| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.1353154938883383, | |
| "eval_steps": 500, | |
| "global_step": 64, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002114304592005286, | |
| "grad_norm": 1.5018059015274048, | |
| "learning_rate": 2.083333333333333e-08, | |
| "logits/chosen": -0.3466828167438507, | |
| "logits/rejected": -0.30099987983703613, | |
| "logps/chosen": -0.9345186948776245, | |
| "logps/rejected": -0.9117153882980347, | |
| "loss": 1.4889, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.869037389755249, | |
| "rewards/margins": -0.04560665041208267, | |
| "rewards/rejected": -1.8234307765960693, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004228609184010572, | |
| "grad_norm": 0.8093975186347961, | |
| "learning_rate": 4.166666666666666e-08, | |
| "logits/chosen": -0.4310421049594879, | |
| "logits/rejected": -0.39132067561149597, | |
| "logps/chosen": -0.8198825716972351, | |
| "logps/rejected": -0.8644211888313293, | |
| "loss": 1.376, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.6397651433944702, | |
| "rewards/margins": 0.08907715976238251, | |
| "rewards/rejected": -1.7288423776626587, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.006342913776015857, | |
| "grad_norm": 0.5377389788627625, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": -0.46692028641700745, | |
| "logits/rejected": -0.4649256467819214, | |
| "logps/chosen": -0.9087910652160645, | |
| "logps/rejected": -0.9648240804672241, | |
| "loss": 1.3404, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.817582130432129, | |
| "rewards/margins": 0.11206617206335068, | |
| "rewards/rejected": -1.9296481609344482, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.008457218368021144, | |
| "grad_norm": 0.3221875727176666, | |
| "learning_rate": 8.333333333333333e-08, | |
| "logits/chosen": -0.416828453540802, | |
| "logits/rejected": -0.3584724962711334, | |
| "logps/chosen": -0.7818898558616638, | |
| "logps/rejected": -0.8170815110206604, | |
| "loss": 1.3806, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -1.5637797117233276, | |
| "rewards/margins": 0.07038339227437973, | |
| "rewards/rejected": -1.6341630220413208, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.010571522960026428, | |
| "grad_norm": 0.64655601978302, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -0.376886248588562, | |
| "logits/rejected": -0.3516141474246979, | |
| "logps/chosen": -0.8814125061035156, | |
| "logps/rejected": -1.0214396715164185, | |
| "loss": 1.2741, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.7628250122070312, | |
| "rewards/margins": 0.28005433082580566, | |
| "rewards/rejected": -2.042879343032837, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.012685827552031714, | |
| "grad_norm": 0.4775894582271576, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": -0.4757865369319916, | |
| "logits/rejected": -0.4498941898345947, | |
| "logps/chosen": -0.8962199687957764, | |
| "logps/rejected": -0.9462199807167053, | |
| "loss": 1.364, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.7924399375915527, | |
| "rewards/margins": 0.10000008344650269, | |
| "rewards/rejected": -1.8924399614334106, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.014800132144037, | |
| "grad_norm": 1.2459568977355957, | |
| "learning_rate": 1.4583333333333335e-07, | |
| "logits/chosen": -0.38895344734191895, | |
| "logits/rejected": -0.38165366649627686, | |
| "logps/chosen": -0.9025766253471375, | |
| "logps/rejected": -0.9465017318725586, | |
| "loss": 1.3898, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.805153250694275, | |
| "rewards/margins": 0.0878501906991005, | |
| "rewards/rejected": -1.8930034637451172, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.016914436736042288, | |
| "grad_norm": 0.6195729374885559, | |
| "learning_rate": 1.6666666666666665e-07, | |
| "logits/chosen": -0.3964853286743164, | |
| "logits/rejected": -0.377862811088562, | |
| "logps/chosen": -0.9054160118103027, | |
| "logps/rejected": -0.9605879187583923, | |
| "loss": 1.3821, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -1.8108320236206055, | |
| "rewards/margins": 0.1103438138961792, | |
| "rewards/rejected": -1.9211758375167847, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.019028741328047574, | |
| "grad_norm": 1.2074137926101685, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": -0.3729037344455719, | |
| "logits/rejected": -0.38143450021743774, | |
| "logps/chosen": -0.9328653216362, | |
| "logps/rejected": -0.9905799627304077, | |
| "loss": 1.3754, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.8657306432724, | |
| "rewards/margins": 0.11542946100234985, | |
| "rewards/rejected": -1.9811599254608154, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.021143045920052856, | |
| "grad_norm": 0.2867220640182495, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -0.4263336658477783, | |
| "logits/rejected": -0.42903271317481995, | |
| "logps/chosen": -0.8979260325431824, | |
| "logps/rejected": -0.9078099727630615, | |
| "loss": 1.4438, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.7958520650863647, | |
| "rewards/margins": 0.019767940044403076, | |
| "rewards/rejected": -1.815619945526123, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.023257350512058142, | |
| "grad_norm": 0.8363026976585388, | |
| "learning_rate": 2.2916666666666663e-07, | |
| "logits/chosen": -0.3374914526939392, | |
| "logits/rejected": -0.32399696111679077, | |
| "logps/chosen": -0.8886098861694336, | |
| "logps/rejected": -0.9484556317329407, | |
| "loss": 1.3422, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.7772197723388672, | |
| "rewards/margins": 0.11969132721424103, | |
| "rewards/rejected": -1.8969112634658813, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.025371655104063428, | |
| "grad_norm": 0.5406804084777832, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -0.42844679951667786, | |
| "logits/rejected": -0.37984615564346313, | |
| "logps/chosen": -0.861629843711853, | |
| "logps/rejected": -0.8968492150306702, | |
| "loss": 1.3922, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.723259687423706, | |
| "rewards/margins": 0.07043875753879547, | |
| "rewards/rejected": -1.7936984300613403, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.027485959696068714, | |
| "grad_norm": 0.9919329285621643, | |
| "learning_rate": 2.708333333333333e-07, | |
| "logits/chosen": -0.36495402455329895, | |
| "logits/rejected": -0.3249490261077881, | |
| "logps/chosen": -0.8502095937728882, | |
| "logps/rejected": -0.8470643758773804, | |
| "loss": 1.4334, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.7004191875457764, | |
| "rewards/margins": -0.006290358491241932, | |
| "rewards/rejected": -1.6941287517547607, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.029600264288074, | |
| "grad_norm": 0.5477162003517151, | |
| "learning_rate": 2.916666666666667e-07, | |
| "logits/chosen": -0.4155704081058502, | |
| "logits/rejected": -0.39535820484161377, | |
| "logps/chosen": -1.0430240631103516, | |
| "logps/rejected": -1.1318373680114746, | |
| "loss": 1.3533, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -2.086048126220703, | |
| "rewards/margins": 0.17762640118598938, | |
| "rewards/rejected": -2.263674736022949, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.031714568880079286, | |
| "grad_norm": 0.26530712842941284, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -0.4810572564601898, | |
| "logits/rejected": -0.42454615235328674, | |
| "logps/chosen": -0.8741041421890259, | |
| "logps/rejected": -0.9494178295135498, | |
| "loss": 1.3655, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.7482082843780518, | |
| "rewards/margins": 0.15062758326530457, | |
| "rewards/rejected": -1.8988356590270996, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.033828873472084575, | |
| "grad_norm": 0.9272629618644714, | |
| "learning_rate": 3.333333333333333e-07, | |
| "logits/chosen": -0.4440098702907562, | |
| "logits/rejected": -0.3930297791957855, | |
| "logps/chosen": -0.8473359942436218, | |
| "logps/rejected": -0.9369213581085205, | |
| "loss": 1.3248, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.6946719884872437, | |
| "rewards/margins": 0.17917080223560333, | |
| "rewards/rejected": -1.873842716217041, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03594317806408986, | |
| "grad_norm": 0.5912418961524963, | |
| "learning_rate": 3.541666666666667e-07, | |
| "logits/chosen": -0.3838099539279938, | |
| "logits/rejected": -0.3507584035396576, | |
| "logps/chosen": -0.8888350129127502, | |
| "logps/rejected": -0.9361770749092102, | |
| "loss": 1.383, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -1.7776700258255005, | |
| "rewards/margins": 0.0946839451789856, | |
| "rewards/rejected": -1.8723541498184204, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03805748265609515, | |
| "grad_norm": 0.6536504030227661, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": -0.3581697940826416, | |
| "logits/rejected": -0.3620460629463196, | |
| "logps/chosen": -0.8519617319107056, | |
| "logps/rejected": -0.9022184610366821, | |
| "loss": 1.3841, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.7039234638214111, | |
| "rewards/margins": 0.10051343590021133, | |
| "rewards/rejected": -1.8044369220733643, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04017178724810043, | |
| "grad_norm": 0.3433632552623749, | |
| "learning_rate": 3.958333333333333e-07, | |
| "logits/chosen": -0.37887442111968994, | |
| "logits/rejected": -0.37543320655822754, | |
| "logps/chosen": -0.9464104175567627, | |
| "logps/rejected": -1.0017329454421997, | |
| "loss": 1.3649, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.8928208351135254, | |
| "rewards/margins": 0.11064518243074417, | |
| "rewards/rejected": -2.0034658908843994, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04228609184010571, | |
| "grad_norm": 0.9764007329940796, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -0.44110679626464844, | |
| "logits/rejected": -0.4280649721622467, | |
| "logps/chosen": -0.9046768546104431, | |
| "logps/rejected": -1.0464633703231812, | |
| "loss": 1.2592, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.8093537092208862, | |
| "rewards/margins": 0.2835729420185089, | |
| "rewards/rejected": -2.0929267406463623, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.044400396432111, | |
| "grad_norm": 1.8563830852508545, | |
| "learning_rate": 4.375e-07, | |
| "logits/chosen": -0.45183491706848145, | |
| "logits/rejected": -0.42935287952423096, | |
| "logps/chosen": -0.9043138027191162, | |
| "logps/rejected": -0.9462392926216125, | |
| "loss": 1.3784, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.8086276054382324, | |
| "rewards/margins": 0.08385094255208969, | |
| "rewards/rejected": -1.892478585243225, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.046514701024116284, | |
| "grad_norm": 1.3473299741744995, | |
| "learning_rate": 4.5833333333333327e-07, | |
| "logits/chosen": -0.37855517864227295, | |
| "logits/rejected": -0.34429043531417847, | |
| "logps/chosen": -0.9284683465957642, | |
| "logps/rejected": -0.9454050064086914, | |
| "loss": 1.4346, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -1.8569366931915283, | |
| "rewards/margins": 0.03387312963604927, | |
| "rewards/rejected": -1.8908100128173828, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04862900561612157, | |
| "grad_norm": 0.940831184387207, | |
| "learning_rate": 4.791666666666667e-07, | |
| "logits/chosen": -0.39172160625457764, | |
| "logits/rejected": -0.3695780634880066, | |
| "logps/chosen": -0.9314202666282654, | |
| "logps/rejected": -1.020229697227478, | |
| "loss": 1.3322, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.8628405332565308, | |
| "rewards/margins": 0.17761869728565216, | |
| "rewards/rejected": -2.040459394454956, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.050743310208126856, | |
| "grad_norm": 0.5783158540725708, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -0.4958629608154297, | |
| "logits/rejected": -0.4257377088069916, | |
| "logps/chosen": -0.9379237294197083, | |
| "logps/rejected": -0.9415461421012878, | |
| "loss": 1.441, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.8758474588394165, | |
| "rewards/margins": 0.0072449808940291405, | |
| "rewards/rejected": -1.8830922842025757, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.052857614800132145, | |
| "grad_norm": 1.4209853410720825, | |
| "learning_rate": 5.208333333333334e-07, | |
| "logits/chosen": -0.36407172679901123, | |
| "logits/rejected": -0.3331725299358368, | |
| "logps/chosen": -0.9192589521408081, | |
| "logps/rejected": -0.9595308899879456, | |
| "loss": 1.3994, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.8385179042816162, | |
| "rewards/margins": 0.080544114112854, | |
| "rewards/rejected": -1.9190617799758911, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05497191939213743, | |
| "grad_norm": 0.6310216188430786, | |
| "learning_rate": 5.416666666666666e-07, | |
| "logits/chosen": -0.41772690415382385, | |
| "logits/rejected": -0.36565953493118286, | |
| "logps/chosen": -0.8052878379821777, | |
| "logps/rejected": -0.8673746585845947, | |
| "loss": 1.3356, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -1.6105756759643555, | |
| "rewards/margins": 0.12417369335889816, | |
| "rewards/rejected": -1.7347493171691895, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05708622398414272, | |
| "grad_norm": 1.2933462858200073, | |
| "learning_rate": 5.625e-07, | |
| "logits/chosen": -0.4482795000076294, | |
| "logits/rejected": -0.39409321546554565, | |
| "logps/chosen": -0.8339261412620544, | |
| "logps/rejected": -0.8675202131271362, | |
| "loss": 1.3739, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -1.6678522825241089, | |
| "rewards/margins": 0.06718815863132477, | |
| "rewards/rejected": -1.7350404262542725, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.059200528576148, | |
| "grad_norm": 0.5808025002479553, | |
| "learning_rate": 5.833333333333334e-07, | |
| "logits/chosen": -0.37116044759750366, | |
| "logits/rejected": -0.3478051722049713, | |
| "logps/chosen": -0.8950318694114685, | |
| "logps/rejected": -0.9756672978401184, | |
| "loss": 1.3505, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.790063738822937, | |
| "rewards/margins": 0.1612708568572998, | |
| "rewards/rejected": -1.9513345956802368, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06131483316815329, | |
| "grad_norm": 1.0569533109664917, | |
| "learning_rate": 6.041666666666666e-07, | |
| "logits/chosen": -0.421148419380188, | |
| "logits/rejected": -0.38443076610565186, | |
| "logps/chosen": -0.8021283745765686, | |
| "logps/rejected": -0.8370179533958435, | |
| "loss": 1.3916, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -1.6042567491531372, | |
| "rewards/margins": 0.06977920234203339, | |
| "rewards/rejected": -1.674035906791687, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.06342913776015857, | |
| "grad_norm": 0.42577147483825684, | |
| "learning_rate": 6.249999999999999e-07, | |
| "logits/chosen": -0.4429818391799927, | |
| "logits/rejected": -0.3524704575538635, | |
| "logps/chosen": -0.8916822671890259, | |
| "logps/rejected": -0.8985542058944702, | |
| "loss": 1.4321, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -1.7833645343780518, | |
| "rewards/margins": 0.01374388113617897, | |
| "rewards/rejected": -1.7971084117889404, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06554344235216386, | |
| "grad_norm": 1.0056904554367065, | |
| "learning_rate": 6.458333333333333e-07, | |
| "logits/chosen": -0.376451700925827, | |
| "logits/rejected": -0.342519074678421, | |
| "logps/chosen": -0.9038617014884949, | |
| "logps/rejected": -0.953092634677887, | |
| "loss": 1.398, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -1.8077234029769897, | |
| "rewards/margins": 0.09846188127994537, | |
| "rewards/rejected": -1.906185269355774, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.06765774694416915, | |
| "grad_norm": 0.5494012236595154, | |
| "learning_rate": 6.666666666666666e-07, | |
| "logits/chosen": -0.3459138870239258, | |
| "logits/rejected": -0.3590989410877228, | |
| "logps/chosen": -0.8274999260902405, | |
| "logps/rejected": -0.8776509761810303, | |
| "loss": 1.363, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.654999852180481, | |
| "rewards/margins": 0.1003020703792572, | |
| "rewards/rejected": -1.7553019523620605, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.06977205153617443, | |
| "grad_norm": 0.693267822265625, | |
| "learning_rate": 6.875e-07, | |
| "logits/chosen": -0.40053680539131165, | |
| "logits/rejected": -0.37323904037475586, | |
| "logps/chosen": -0.8255244493484497, | |
| "logps/rejected": -0.8658804893493652, | |
| "loss": 1.3712, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -1.6510488986968994, | |
| "rewards/margins": 0.08071210980415344, | |
| "rewards/rejected": -1.7317609786987305, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.07188635612817972, | |
| "grad_norm": 2.213238000869751, | |
| "learning_rate": 7.083333333333334e-07, | |
| "logits/chosen": -0.40097948908805847, | |
| "logits/rejected": -0.38190510869026184, | |
| "logps/chosen": -0.9122671484947205, | |
| "logps/rejected": -0.9549552798271179, | |
| "loss": 1.36, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -1.824534296989441, | |
| "rewards/margins": 0.0853763073682785, | |
| "rewards/rejected": -1.9099105596542358, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.074000660720185, | |
| "grad_norm": 0.6859830021858215, | |
| "learning_rate": 7.291666666666666e-07, | |
| "logits/chosen": -0.42501094937324524, | |
| "logits/rejected": -0.42549416422843933, | |
| "logps/chosen": -1.0008373260498047, | |
| "logps/rejected": -1.1157118082046509, | |
| "loss": 1.3294, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -2.0016746520996094, | |
| "rewards/margins": 0.22974897921085358, | |
| "rewards/rejected": -2.2314236164093018, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0761149653121903, | |
| "grad_norm": 0.6468721628189087, | |
| "learning_rate": 7.5e-07, | |
| "logits/chosen": -0.36494994163513184, | |
| "logits/rejected": -0.30433908104896545, | |
| "logps/chosen": -0.9062094688415527, | |
| "logps/rejected": -0.920263409614563, | |
| "loss": 1.4312, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -1.8124189376831055, | |
| "rewards/margins": 0.02810765616595745, | |
| "rewards/rejected": -1.840526819229126, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.07822926990419557, | |
| "grad_norm": 0.5085556507110596, | |
| "learning_rate": 7.708333333333333e-07, | |
| "logits/chosen": -0.4677881598472595, | |
| "logits/rejected": -0.456132709980011, | |
| "logps/chosen": -1.0101865530014038, | |
| "logps/rejected": -1.0429682731628418, | |
| "loss": 1.4132, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -2.0203731060028076, | |
| "rewards/margins": 0.06556359678506851, | |
| "rewards/rejected": -2.0859365463256836, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08034357449620086, | |
| "grad_norm": 0.23813335597515106, | |
| "learning_rate": 7.916666666666666e-07, | |
| "logits/chosen": -0.3991190791130066, | |
| "logits/rejected": -0.3664044141769409, | |
| "logps/chosen": -0.9578174352645874, | |
| "logps/rejected": -0.9229263067245483, | |
| "loss": 1.4824, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": -1.9156348705291748, | |
| "rewards/margins": -0.06978224962949753, | |
| "rewards/rejected": -1.8458526134490967, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08245787908820615, | |
| "grad_norm": 0.587037980556488, | |
| "learning_rate": 8.125e-07, | |
| "logits/chosen": -0.37554049491882324, | |
| "logits/rejected": -0.36305734515190125, | |
| "logps/chosen": -0.8503091931343079, | |
| "logps/rejected": -0.864615261554718, | |
| "loss": 1.4086, | |
| "rewards/accuracies": 0.4296875, | |
| "rewards/chosen": -1.7006183862686157, | |
| "rewards/margins": 0.028611989691853523, | |
| "rewards/rejected": -1.729230523109436, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.08457218368021142, | |
| "grad_norm": 0.4172501862049103, | |
| "learning_rate": 8.333333333333333e-07, | |
| "logits/chosen": -0.4405443072319031, | |
| "logits/rejected": -0.41723060607910156, | |
| "logps/chosen": -0.8502858877182007, | |
| "logps/rejected": -0.9114271402359009, | |
| "loss": 1.3446, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.7005717754364014, | |
| "rewards/margins": 0.12228240817785263, | |
| "rewards/rejected": -1.8228542804718018, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08668648827221671, | |
| "grad_norm": 0.9275372624397278, | |
| "learning_rate": 8.541666666666666e-07, | |
| "logits/chosen": -0.4200601577758789, | |
| "logits/rejected": -0.3478623628616333, | |
| "logps/chosen": -0.892408013343811, | |
| "logps/rejected": -0.9276402592658997, | |
| "loss": 1.3887, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -1.784816026687622, | |
| "rewards/margins": 0.07046431303024292, | |
| "rewards/rejected": -1.8552805185317993, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.088800792864222, | |
| "grad_norm": 0.7317383289337158, | |
| "learning_rate": 8.75e-07, | |
| "logits/chosen": -0.37675267457962036, | |
| "logits/rejected": -0.33540332317352295, | |
| "logps/chosen": -0.7866061925888062, | |
| "logps/rejected": -0.824250340461731, | |
| "loss": 1.3837, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.5732123851776123, | |
| "rewards/margins": 0.07528844475746155, | |
| "rewards/rejected": -1.648500680923462, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09091509745622729, | |
| "grad_norm": 0.9452736973762512, | |
| "learning_rate": 8.958333333333334e-07, | |
| "logits/chosen": -0.4662383198738098, | |
| "logits/rejected": -0.4447881579399109, | |
| "logps/chosen": -0.9490666389465332, | |
| "logps/rejected": -1.0112388134002686, | |
| "loss": 1.3412, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -1.8981332778930664, | |
| "rewards/margins": 0.12434446066617966, | |
| "rewards/rejected": -2.022477626800537, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.09302940204823257, | |
| "grad_norm": 0.2848323881626129, | |
| "learning_rate": 9.166666666666665e-07, | |
| "logits/chosen": -0.41404005885124207, | |
| "logits/rejected": -0.3944583535194397, | |
| "logps/chosen": -0.8224930167198181, | |
| "logps/rejected": -0.8416361808776855, | |
| "loss": 1.4027, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -1.6449860334396362, | |
| "rewards/margins": 0.038286346942186356, | |
| "rewards/rejected": -1.683272361755371, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.09514370664023786, | |
| "grad_norm": 0.7165678143501282, | |
| "learning_rate": 9.374999999999999e-07, | |
| "logits/chosen": -0.40475326776504517, | |
| "logits/rejected": -0.3559921383857727, | |
| "logps/chosen": -0.8070214986801147, | |
| "logps/rejected": -0.8993593454360962, | |
| "loss": 1.3148, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.6140429973602295, | |
| "rewards/margins": 0.18467575311660767, | |
| "rewards/rejected": -1.7987186908721924, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.09725801123224315, | |
| "grad_norm": 0.4779021739959717, | |
| "learning_rate": 9.583333333333334e-07, | |
| "logits/chosen": -0.4171525835990906, | |
| "logits/rejected": -0.42166149616241455, | |
| "logps/chosen": -0.7872560024261475, | |
| "logps/rejected": -0.8496187925338745, | |
| "loss": 1.3356, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.574512004852295, | |
| "rewards/margins": 0.12472567707300186, | |
| "rewards/rejected": -1.699237585067749, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.09937231582424844, | |
| "grad_norm": 0.7870219349861145, | |
| "learning_rate": 9.791666666666667e-07, | |
| "logits/chosen": -0.3734116554260254, | |
| "logits/rejected": -0.32778748869895935, | |
| "logps/chosen": -0.7842286825180054, | |
| "logps/rejected": -0.8161548972129822, | |
| "loss": 1.3647, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.5684573650360107, | |
| "rewards/margins": 0.06385258585214615, | |
| "rewards/rejected": -1.6323097944259644, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.10148662041625371, | |
| "grad_norm": 0.2597256600856781, | |
| "learning_rate": 1e-06, | |
| "logits/chosen": -0.4355677664279938, | |
| "logits/rejected": -0.38983187079429626, | |
| "logps/chosen": -0.8787693977355957, | |
| "logps/rejected": -0.9383041262626648, | |
| "loss": 1.35, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.7575387954711914, | |
| "rewards/margins": 0.11906948685646057, | |
| "rewards/rejected": -1.8766082525253296, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.103600925008259, | |
| "grad_norm": 0.9942799210548401, | |
| "learning_rate": 9.999862751990697e-07, | |
| "logits/chosen": -0.4244321882724762, | |
| "logits/rejected": -0.4366786777973175, | |
| "logps/chosen": -0.7910157442092896, | |
| "logps/rejected": -0.8630884885787964, | |
| "loss": 1.3166, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.582031488418579, | |
| "rewards/margins": 0.14414538443088531, | |
| "rewards/rejected": -1.7261769771575928, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.10571522960026429, | |
| "grad_norm": 0.5333903431892395, | |
| "learning_rate": 9.999451015497595e-07, | |
| "logits/chosen": -0.389942467212677, | |
| "logits/rejected": -0.36674585938453674, | |
| "logps/chosen": -0.7312074899673462, | |
| "logps/rejected": -0.7289648652076721, | |
| "loss": 1.4225, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -1.4624149799346924, | |
| "rewards/margins": -0.004485193639993668, | |
| "rewards/rejected": -1.4579297304153442, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10782953419226958, | |
| "grad_norm": 0.5712242722511292, | |
| "learning_rate": 9.9987648131247e-07, | |
| "logits/chosen": -0.4622853994369507, | |
| "logits/rejected": -0.3728552460670471, | |
| "logps/chosen": -0.8764299750328064, | |
| "logps/rejected": -0.869678795337677, | |
| "loss": 1.4542, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.7528599500656128, | |
| "rewards/margins": -0.013502337038516998, | |
| "rewards/rejected": -1.739357590675354, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.10994383878427486, | |
| "grad_norm": 0.2586441934108734, | |
| "learning_rate": 9.99780418254397e-07, | |
| "logits/chosen": -0.37249019742012024, | |
| "logits/rejected": -0.3998304605484009, | |
| "logps/chosen": -0.8435611724853516, | |
| "logps/rejected": -0.9359882473945618, | |
| "loss": 1.3057, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.6871223449707031, | |
| "rewards/margins": 0.18485431373119354, | |
| "rewards/rejected": -1.8719764947891235, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.11205814337628014, | |
| "grad_norm": 1.0829113721847534, | |
| "learning_rate": 9.996569176493268e-07, | |
| "logits/chosen": -0.47697725892066956, | |
| "logits/rejected": -0.4208195209503174, | |
| "logps/chosen": -0.8014968037605286, | |
| "logps/rejected": -0.8703804612159729, | |
| "loss": 1.3523, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.6029936075210571, | |
| "rewards/margins": 0.1377674788236618, | |
| "rewards/rejected": -1.7407609224319458, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.11417244796828543, | |
| "grad_norm": 0.5523208379745483, | |
| "learning_rate": 9.995059862773438e-07, | |
| "logits/chosen": -0.40533363819122314, | |
| "logits/rejected": -0.36801978945732117, | |
| "logps/chosen": -0.7641825675964355, | |
| "logps/rejected": -0.8168596029281616, | |
| "loss": 1.3692, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.528365135192871, | |
| "rewards/margins": 0.10535416752099991, | |
| "rewards/rejected": -1.6337192058563232, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.11628675256029072, | |
| "grad_norm": 0.614101767539978, | |
| "learning_rate": 9.993276324244605e-07, | |
| "logits/chosen": -0.4476906955242157, | |
| "logits/rejected": -0.40396648645401, | |
| "logps/chosen": -0.8706808090209961, | |
| "logps/rejected": -0.9221430420875549, | |
| "loss": 1.3787, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.7413616180419922, | |
| "rewards/margins": 0.10292442888021469, | |
| "rewards/rejected": -1.8442860841751099, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.118401057152296, | |
| "grad_norm": 0.3428778052330017, | |
| "learning_rate": 9.991218658821608e-07, | |
| "logits/chosen": -0.31709593534469604, | |
| "logits/rejected": -0.2760937213897705, | |
| "logps/chosen": -0.842248797416687, | |
| "logps/rejected": -0.8068034648895264, | |
| "loss": 1.498, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -1.684497594833374, | |
| "rewards/margins": -0.07089066505432129, | |
| "rewards/rejected": -1.6136069297790527, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.12051536174430129, | |
| "grad_norm": 0.6877723932266235, | |
| "learning_rate": 9.988886979468643e-07, | |
| "logits/chosen": -0.41800016164779663, | |
| "logits/rejected": -0.4011584222316742, | |
| "logps/chosen": -0.7845420837402344, | |
| "logps/rejected": -0.834447979927063, | |
| "loss": 1.3491, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.5690841674804688, | |
| "rewards/margins": 0.09981165081262589, | |
| "rewards/rejected": -1.668895959854126, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.12262966633630658, | |
| "grad_norm": 0.9649701714515686, | |
| "learning_rate": 9.98628141419305e-07, | |
| "logits/chosen": -0.4253537058830261, | |
| "logits/rejected": -0.4305458962917328, | |
| "logps/chosen": -0.86476731300354, | |
| "logps/rejected": -0.9080386161804199, | |
| "loss": 1.3639, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.72953462600708, | |
| "rewards/margins": 0.08654248714447021, | |
| "rewards/rejected": -1.8160772323608398, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.12474397092831185, | |
| "grad_norm": 1.3779780864715576, | |
| "learning_rate": 9.98340210603829e-07, | |
| "logits/chosen": -0.39970022439956665, | |
| "logits/rejected": -0.441428005695343, | |
| "logps/chosen": -0.8662775158882141, | |
| "logps/rejected": -0.9646260738372803, | |
| "loss": 1.3001, | |
| "rewards/accuracies": 0.6484375, | |
| "rewards/chosen": -1.7325550317764282, | |
| "rewards/margins": 0.19669723510742188, | |
| "rewards/rejected": -1.9292521476745605, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.12685827552031714, | |
| "grad_norm": 0.5366966724395752, | |
| "learning_rate": 9.980249213076084e-07, | |
| "logits/chosen": -0.37770116329193115, | |
| "logits/rejected": -0.35231757164001465, | |
| "logps/chosen": -0.8165755867958069, | |
| "logps/rejected": -0.8619179129600525, | |
| "loss": 1.3699, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -1.6331511735916138, | |
| "rewards/margins": 0.09068439900875092, | |
| "rewards/rejected": -1.723835825920105, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12897258011232243, | |
| "grad_norm": 0.36810922622680664, | |
| "learning_rate": 9.976822908397748e-07, | |
| "logits/chosen": -0.4224976897239685, | |
| "logits/rejected": -0.41758257150650024, | |
| "logps/chosen": -0.8445641994476318, | |
| "logps/rejected": -0.9393664598464966, | |
| "loss": 1.3193, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -1.6891283988952637, | |
| "rewards/margins": 0.18960458040237427, | |
| "rewards/rejected": -1.8787329196929932, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.13108688470432772, | |
| "grad_norm": 0.6838279366493225, | |
| "learning_rate": 9.97312338010468e-07, | |
| "logits/chosen": -0.4168627858161926, | |
| "logits/rejected": -0.36115381121635437, | |
| "logps/chosen": -0.8370552659034729, | |
| "logps/rejected": -0.8352169394493103, | |
| "loss": 1.4284, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.6741105318069458, | |
| "rewards/margins": -0.0036766715347766876, | |
| "rewards/rejected": -1.6704338788986206, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.133201189296333, | |
| "grad_norm": 0.39330533146858215, | |
| "learning_rate": 9.969150831298037e-07, | |
| "logits/chosen": -0.4558233618736267, | |
| "logits/rejected": -0.4025765061378479, | |
| "logps/chosen": -0.826255738735199, | |
| "logps/rejected": -0.894213080406189, | |
| "loss": 1.3485, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.652511477470398, | |
| "rewards/margins": 0.13591471314430237, | |
| "rewards/rejected": -1.788426160812378, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1353154938883383, | |
| "grad_norm": 0.6055929660797119, | |
| "learning_rate": 9.964905480067584e-07, | |
| "logits/chosen": -0.459463506937027, | |
| "logits/rejected": -0.42943331599235535, | |
| "logps/chosen": -0.7901928424835205, | |
| "logps/rejected": -0.7964221239089966, | |
| "loss": 1.4057, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -1.580385684967041, | |
| "rewards/margins": 0.012458762153983116, | |
| "rewards/rejected": -1.5928442478179932, | |
| "step": 64 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 472, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 64, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |