| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9988623435722411, | |
| "eval_steps": 10000000, | |
| "global_step": 439, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 113.4577168560686, | |
| "learning_rate": 2.2727272727272727e-09, | |
| "logits/chosen": -1.6768856048583984, | |
| "logits/rejected": -1.7259055376052856, | |
| "logps/chosen": -1.2793102264404297, | |
| "logps/rejected": -1.2162058353424072, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 129.24988403709358, | |
| "learning_rate": 2.2727272727272725e-08, | |
| "logits/chosen": -1.7028687000274658, | |
| "logits/rejected": -1.6680525541305542, | |
| "logps/chosen": -1.2131904363632202, | |
| "logps/rejected": -1.2204842567443848, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.4652777910232544, | |
| "rewards/chosen": 0.0011519736144691706, | |
| "rewards/margins": -0.001473029493354261, | |
| "rewards/rejected": 0.0026250029914081097, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 128.90531745075424, | |
| "learning_rate": 4.545454545454545e-08, | |
| "logits/chosen": -1.7795600891113281, | |
| "logits/rejected": -1.7348560094833374, | |
| "logps/chosen": -1.1446261405944824, | |
| "logps/rejected": -1.1849511861801147, | |
| "loss": 0.6903, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.01403624564409256, | |
| "rewards/margins": 0.0008796676993370056, | |
| "rewards/rejected": -0.014915913343429565, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 130.88686983721493, | |
| "learning_rate": 6.818181818181817e-08, | |
| "logits/chosen": -1.7440669536590576, | |
| "logits/rejected": -1.6752160787582397, | |
| "logps/chosen": -1.194157600402832, | |
| "logps/rejected": -1.246543526649475, | |
| "loss": 0.6755, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.11256156116724014, | |
| "rewards/margins": 0.05156043916940689, | |
| "rewards/rejected": -0.16412200033664703, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 99.69346579706567, | |
| "learning_rate": 9.09090909090909e-08, | |
| "logits/chosen": -1.7316129207611084, | |
| "logits/rejected": -1.6658855676651, | |
| "logps/chosen": -1.2354015111923218, | |
| "logps/rejected": -1.3086662292480469, | |
| "loss": 0.6598, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.28453049063682556, | |
| "rewards/margins": 0.20941098034381866, | |
| "rewards/rejected": -0.49394145607948303, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 94.76247831424028, | |
| "learning_rate": 9.994307990108962e-08, | |
| "logits/chosen": -1.6965711116790771, | |
| "logits/rejected": -1.6329838037490845, | |
| "logps/chosen": -1.254882574081421, | |
| "logps/rejected": -1.3045636415481567, | |
| "loss": 0.6257, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.4679547846317291, | |
| "rewards/margins": 0.29588648676872253, | |
| "rewards/rejected": -0.7638412714004517, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 134.02849477110337, | |
| "learning_rate": 9.959570405988094e-08, | |
| "logits/chosen": -1.7339792251586914, | |
| "logits/rejected": -1.6545541286468506, | |
| "logps/chosen": -1.1747747659683228, | |
| "logps/rejected": -1.2515872716903687, | |
| "loss": 0.628, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.6567646265029907, | |
| "rewards/margins": 0.2488430291414261, | |
| "rewards/rejected": -0.9056077003479004, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 81.9825685637945, | |
| "learning_rate": 9.893476820924666e-08, | |
| "logits/chosen": -1.8203418254852295, | |
| "logits/rejected": -1.7378566265106201, | |
| "logps/chosen": -1.2860430479049683, | |
| "logps/rejected": -1.3670638799667358, | |
| "loss": 0.5884, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.732010006904602, | |
| "rewards/margins": 0.3435601592063904, | |
| "rewards/rejected": -1.0755703449249268, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 111.43014849080478, | |
| "learning_rate": 9.796445099843647e-08, | |
| "logits/chosen": -1.8093068599700928, | |
| "logits/rejected": -1.7265288829803467, | |
| "logps/chosen": -1.2918837070465088, | |
| "logps/rejected": -1.3895039558410645, | |
| "loss": 0.6282, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.8491527438163757, | |
| "rewards/margins": 0.3709031641483307, | |
| "rewards/rejected": -1.2200558185577393, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 99.11416934571663, | |
| "learning_rate": 9.669088708527066e-08, | |
| "logits/chosen": -1.756994605064392, | |
| "logits/rejected": -1.6932170391082764, | |
| "logps/chosen": -1.3391311168670654, | |
| "logps/rejected": -1.3889451026916504, | |
| "loss": 0.5898, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0343393087387085, | |
| "rewards/margins": 0.4276931881904602, | |
| "rewards/rejected": -1.462032437324524, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 108.94100525225643, | |
| "learning_rate": 9.512212835085849e-08, | |
| "logits/chosen": -1.7921009063720703, | |
| "logits/rejected": -1.7081331014633179, | |
| "logps/chosen": -1.2848669290542603, | |
| "logps/rejected": -1.379817247390747, | |
| "loss": 0.5831, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.1641167402267456, | |
| "rewards/margins": 0.5224048495292664, | |
| "rewards/rejected": -1.6865215301513672, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 94.58328257186776, | |
| "learning_rate": 9.326809299301306e-08, | |
| "logits/chosen": -1.7934032678604126, | |
| "logits/rejected": -1.692413568496704, | |
| "logps/chosen": -1.3193799257278442, | |
| "logps/rejected": -1.4346280097961426, | |
| "loss": 0.5776, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1873786449432373, | |
| "rewards/margins": 0.6364501714706421, | |
| "rewards/rejected": -1.8238286972045898, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 123.65835816529943, | |
| "learning_rate": 9.114050282021158e-08, | |
| "logits/chosen": -1.7867753505706787, | |
| "logits/rejected": -1.7270488739013672, | |
| "logps/chosen": -1.2704253196716309, | |
| "logps/rejected": -1.373067855834961, | |
| "loss": 0.5626, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.228627324104309, | |
| "rewards/margins": 0.5525014400482178, | |
| "rewards/rejected": -1.7811288833618164, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 95.55607076940454, | |
| "learning_rate": 8.875280914254802e-08, | |
| "logits/chosen": -1.7855722904205322, | |
| "logits/rejected": -1.6952784061431885, | |
| "logps/chosen": -1.3185499906539917, | |
| "logps/rejected": -1.4162604808807373, | |
| "loss": 0.5666, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.2571992874145508, | |
| "rewards/margins": 0.6928548812866211, | |
| "rewards/rejected": -1.9500541687011719, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 78.06311583858023, | |
| "learning_rate": 8.612010772821971e-08, | |
| "logits/chosen": -1.8129308223724365, | |
| "logits/rejected": -1.7682344913482666, | |
| "logps/chosen": -1.3229562044143677, | |
| "logps/rejected": -1.361987829208374, | |
| "loss": 0.5656, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.0907716751098633, | |
| "rewards/margins": 0.6394159197807312, | |
| "rewards/rejected": -1.7301876544952393, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 78.88051547861711, | |
| "learning_rate": 8.325904336322055e-08, | |
| "logits/chosen": -1.7850853204727173, | |
| "logits/rejected": -1.7291721105575562, | |
| "logps/chosen": -1.2866876125335693, | |
| "logps/rejected": -1.402489423751831, | |
| "loss": 0.5657, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1932275295257568, | |
| "rewards/margins": 0.6862825155258179, | |
| "rewards/rejected": -1.8795099258422852, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 80.61975465340338, | |
| "learning_rate": 8.01877046176447e-08, | |
| "logits/chosen": -1.7330862283706665, | |
| "logits/rejected": -1.6600639820098877, | |
| "logps/chosen": -1.312774419784546, | |
| "logps/rejected": -1.419638991355896, | |
| "loss": 0.566, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.3447433710098267, | |
| "rewards/margins": 0.5547876358032227, | |
| "rewards/rejected": -1.8995310068130493, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 78.96860347790852, | |
| "learning_rate": 7.692550948392249e-08, | |
| "logits/chosen": -1.7879507541656494, | |
| "logits/rejected": -1.7219922542572021, | |
| "logps/chosen": -1.3235785961151123, | |
| "logps/rejected": -1.3983043432235718, | |
| "loss": 0.5589, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1332305669784546, | |
| "rewards/margins": 0.6428099870681763, | |
| "rewards/rejected": -1.7760404348373413, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 91.20361542856642, | |
| "learning_rate": 7.349308261002021e-08, | |
| "logits/chosen": -1.744236707687378, | |
| "logits/rejected": -1.685158133506775, | |
| "logps/chosen": -1.3169732093811035, | |
| "logps/rejected": -1.422654390335083, | |
| "loss": 0.5541, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.1988707780838013, | |
| "rewards/margins": 0.5603083372116089, | |
| "rewards/rejected": -1.7591791152954102, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 85.52532511636508, | |
| "learning_rate": 6.991212490377531e-08, | |
| "logits/chosen": -1.7989540100097656, | |
| "logits/rejected": -1.7449232339859009, | |
| "logps/chosen": -1.341355562210083, | |
| "logps/rejected": -1.4322634935379028, | |
| "loss": 0.528, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -1.112524390220642, | |
| "rewards/margins": 0.6936608552932739, | |
| "rewards/rejected": -1.8061851263046265, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 103.62294205062797, | |
| "learning_rate": 6.620527633276978e-08, | |
| "logits/chosen": -1.7379404306411743, | |
| "logits/rejected": -1.6623157262802124, | |
| "logps/chosen": -1.307138204574585, | |
| "logps/rejected": -1.4737988710403442, | |
| "loss": 0.54, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -1.126852035522461, | |
| "rewards/margins": 0.7987160682678223, | |
| "rewards/rejected": -1.9255679845809937, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 97.12306479641116, | |
| "learning_rate": 6.239597278716581e-08, | |
| "logits/chosen": -1.8088064193725586, | |
| "logits/rejected": -1.74566650390625, | |
| "logps/chosen": -1.3601690530776978, | |
| "logps/rejected": -1.421120285987854, | |
| "loss": 0.5357, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.1483900547027588, | |
| "rewards/margins": 0.8161466717720032, | |
| "rewards/rejected": -1.9645369052886963, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 73.94190937238754, | |
| "learning_rate": 5.8508297910462456e-08, | |
| "logits/chosen": -1.75725519657135, | |
| "logits/rejected": -1.6749064922332764, | |
| "logps/chosen": -1.2782678604125977, | |
| "logps/rejected": -1.4215964078903198, | |
| "loss": 0.5266, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -1.147524356842041, | |
| "rewards/margins": 0.8555541038513184, | |
| "rewards/rejected": -2.0030786991119385, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 102.01553665817238, | |
| "learning_rate": 5.456683083494731e-08, | |
| "logits/chosen": -1.7501509189605713, | |
| "logits/rejected": -1.7102434635162354, | |
| "logps/chosen": -1.2995601892471313, | |
| "logps/rejected": -1.3736200332641602, | |
| "loss": 0.5541, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.057513952255249, | |
| "rewards/margins": 0.5031803846359253, | |
| "rewards/rejected": -1.5606944561004639, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 111.25877684367076, | |
| "learning_rate": 5.059649078450834e-08, | |
| "logits/chosen": -1.7446762323379517, | |
| "logits/rejected": -1.6967140436172485, | |
| "logps/chosen": -1.2858344316482544, | |
| "logps/rejected": -1.407447338104248, | |
| "loss": 0.5275, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.174809217453003, | |
| "rewards/margins": 0.595741868019104, | |
| "rewards/rejected": -1.770551085472107, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 89.74032522739078, | |
| "learning_rate": 4.6622379527277186e-08, | |
| "logits/chosen": -1.7489802837371826, | |
| "logits/rejected": -1.6957324743270874, | |
| "logps/chosen": -1.262298345565796, | |
| "logps/rejected": -1.361208438873291, | |
| "loss": 0.5316, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2340971231460571, | |
| "rewards/margins": 0.6583037376403809, | |
| "rewards/rejected": -1.8924009799957275, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 75.51753806333156, | |
| "learning_rate": 4.26696226741691e-08, | |
| "logits/chosen": -1.7651500701904297, | |
| "logits/rejected": -1.6915719509124756, | |
| "logps/chosen": -1.3747735023498535, | |
| "logps/rejected": -1.4673488140106201, | |
| "loss": 0.5304, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.4194996356964111, | |
| "rewards/margins": 0.7260831594467163, | |
| "rewards/rejected": -2.145582914352417, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 81.81714896589709, | |
| "learning_rate": 3.876321082668098e-08, | |
| "logits/chosen": -1.8280465602874756, | |
| "logits/rejected": -1.7529118061065674, | |
| "logps/chosen": -1.3470970392227173, | |
| "logps/rejected": -1.435733437538147, | |
| "loss": 0.5229, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.0860626697540283, | |
| "rewards/margins": 0.7932750582695007, | |
| "rewards/rejected": -1.8793376684188843, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 88.74965472733419, | |
| "learning_rate": 3.492784157826244e-08, | |
| "logits/chosen": -1.7529668807983398, | |
| "logits/rejected": -1.6530554294586182, | |
| "logps/chosen": -1.3322269916534424, | |
| "logps/rejected": -1.4489867687225342, | |
| "loss": 0.5263, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.0935055017471313, | |
| "rewards/margins": 0.8137003779411316, | |
| "rewards/rejected": -1.9072058200836182, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 76.0414974261225, | |
| "learning_rate": 3.118776336817812e-08, | |
| "logits/chosen": -1.7995818853378296, | |
| "logits/rejected": -1.7317774295806885, | |
| "logps/chosen": -1.3204050064086914, | |
| "logps/rejected": -1.4101346731185913, | |
| "loss": 0.5231, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.1715632677078247, | |
| "rewards/margins": 0.8251334428787231, | |
| "rewards/rejected": -1.9966968297958374, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 87.86932907500237, | |
| "learning_rate": 2.7566622175067443e-08, | |
| "logits/chosen": -1.7801201343536377, | |
| "logits/rejected": -1.7111116647720337, | |
| "logps/chosen": -1.3444006443023682, | |
| "logps/rejected": -1.494860053062439, | |
| "loss": 0.5308, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.3581712245941162, | |
| "rewards/margins": 0.7247291803359985, | |
| "rewards/rejected": -2.082900285720825, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 101.39189225311264, | |
| "learning_rate": 2.408731201945432e-08, | |
| "logits/chosen": -1.77533757686615, | |
| "logits/rejected": -1.7264738082885742, | |
| "logps/chosen": -1.313521146774292, | |
| "logps/rejected": -1.4012606143951416, | |
| "loss": 0.5179, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -1.1962579488754272, | |
| "rewards/margins": 0.6758478283882141, | |
| "rewards/rejected": -1.8721058368682861, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 97.55308734787654, | |
| "learning_rate": 2.0771830220378112e-08, | |
| "logits/chosen": -1.7309653759002686, | |
| "logits/rejected": -1.673135757446289, | |
| "logps/chosen": -1.3468959331512451, | |
| "logps/rejected": -1.405505895614624, | |
| "loss": 0.5257, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.240114450454712, | |
| "rewards/margins": 0.6646026968955994, | |
| "rewards/rejected": -1.9047170877456665, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 79.56679512217426, | |
| "learning_rate": 1.7641138321260257e-08, | |
| "logits/chosen": -1.7662233114242554, | |
| "logits/rejected": -1.6911147832870483, | |
| "logps/chosen": -1.3011645078659058, | |
| "logps/rejected": -1.418474555015564, | |
| "loss": 0.5175, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -1.1730973720550537, | |
| "rewards/margins": 0.8977500796318054, | |
| "rewards/rejected": -2.070847511291504, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 78.95443991883164, | |
| "learning_rate": 1.4715029564277793e-08, | |
| "logits/chosen": -1.8377494812011719, | |
| "logits/rejected": -1.785035490989685, | |
| "logps/chosen": -1.3052855730056763, | |
| "logps/rejected": -1.4216973781585693, | |
| "loss": 0.535, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.0062506198883057, | |
| "rewards/margins": 0.7934825420379639, | |
| "rewards/rejected": -1.7997331619262695, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 85.7404253023672, | |
| "learning_rate": 1.2012003751113343e-08, | |
| "logits/chosen": -1.8215789794921875, | |
| "logits/rejected": -1.7556202411651611, | |
| "logps/chosen": -1.2944856882095337, | |
| "logps/rejected": -1.427002191543579, | |
| "loss": 0.5037, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -1.3231576681137085, | |
| "rewards/margins": 0.8232595324516296, | |
| "rewards/rejected": -2.1464171409606934, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 81.80824877374742, | |
| "learning_rate": 9.549150281252633e-09, | |
| "logits/chosen": -1.7837785482406616, | |
| "logits/rejected": -1.7289537191390991, | |
| "logps/chosen": -1.2906091213226318, | |
| "logps/rejected": -1.4211328029632568, | |
| "loss": 0.5085, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.1847398281097412, | |
| "rewards/margins": 0.7879296541213989, | |
| "rewards/rejected": -1.9726696014404297, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 76.45542191386993, | |
| "learning_rate": 7.3420401072985306e-09, | |
| "logits/chosen": -1.8155953884124756, | |
| "logits/rejected": -1.7616589069366455, | |
| "logps/chosen": -1.3416122198104858, | |
| "logps/rejected": -1.4825923442840576, | |
| "loss": 0.512, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.2699860334396362, | |
| "rewards/margins": 0.8351644277572632, | |
| "rewards/rejected": -2.1051506996154785, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 100.61500636796463, | |
| "learning_rate": 5.404627290395369e-09, | |
| "logits/chosen": -1.7809003591537476, | |
| "logits/rejected": -1.7117547988891602, | |
| "logps/chosen": -1.299863338470459, | |
| "logps/rejected": -1.4180821180343628, | |
| "loss": 0.5139, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.1025749444961548, | |
| "rewards/margins": 0.7822272777557373, | |
| "rewards/rejected": -1.884802222251892, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 143.00077069598655, | |
| "learning_rate": 3.74916077816162e-09, | |
| "logits/chosen": -1.7901942729949951, | |
| "logits/rejected": -1.7254260778427124, | |
| "logps/chosen": -1.2824052572250366, | |
| "logps/rejected": -1.3659498691558838, | |
| "loss": 0.5278, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.326103925704956, | |
| "rewards/margins": 0.6741579174995422, | |
| "rewards/rejected": -2.0002617835998535, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 86.78461219124101, | |
| "learning_rate": 2.386106962899165e-09, | |
| "logits/chosen": -1.7145483493804932, | |
| "logits/rejected": -1.6393749713897705, | |
| "logps/chosen": -1.3581172227859497, | |
| "logps/rejected": -1.4866178035736084, | |
| "loss": 0.5223, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.3169643878936768, | |
| "rewards/margins": 0.7216087579727173, | |
| "rewards/rejected": -2.0385732650756836, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 97.03178172617403, | |
| "learning_rate": 1.3240835096913706e-09, | |
| "logits/chosen": -1.7489525079727173, | |
| "logits/rejected": -1.6416590213775635, | |
| "logps/chosen": -1.2802146673202515, | |
| "logps/rejected": -1.4399700164794922, | |
| "loss": 0.5334, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.086639404296875, | |
| "rewards/margins": 0.9034107327461243, | |
| "rewards/rejected": -1.9900500774383545, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 177.7844461853396, | |
| "learning_rate": 5.698048727497462e-10, | |
| "logits/chosen": -1.7718130350112915, | |
| "logits/rejected": -1.6983397006988525, | |
| "logps/chosen": -1.3178266286849976, | |
| "logps/rejected": -1.4721015691757202, | |
| "loss": 0.5219, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -1.2652876377105713, | |
| "rewards/margins": 0.8731800317764282, | |
| "rewards/rejected": -2.138467788696289, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 191.4766363479098, | |
| "learning_rate": 1.2803984447259387e-10, | |
| "logits/chosen": -1.7818502187728882, | |
| "logits/rejected": -1.721131682395935, | |
| "logps/chosen": -1.3105113506317139, | |
| "logps/rejected": -1.4450818300247192, | |
| "loss": 0.5113, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.2817203998565674, | |
| "rewards/margins": 0.869175910949707, | |
| "rewards/rejected": -2.1508963108062744, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 439, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5564739256078942, | |
| "train_runtime": 6859.6403, | |
| "train_samples_per_second": 8.198, | |
| "train_steps_per_second": 0.064 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 439, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |