| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 385, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.282051282051282e-07, | |
| "logits/chosen": -1.7278180122375488, | |
| "logits/rejected": -1.7377450466156006, | |
| "logps/chosen": -29.553977966308594, | |
| "logps/rejected": -42.813133239746094, | |
| "loss": 0.5, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.282051282051282e-06, | |
| "logits/chosen": -1.8663586378097534, | |
| "logits/rejected": -1.8706679344177246, | |
| "logps/chosen": -36.9964485168457, | |
| "logps/rejected": -33.65947723388672, | |
| "loss": 0.4966, | |
| "rewards/accuracies": 0.5277777910232544, | |
| "rewards/chosen": 0.005075507797300816, | |
| "rewards/margins": 0.019778331741690636, | |
| "rewards/rejected": -0.014702823013067245, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.564102564102564e-06, | |
| "logits/chosen": -1.9970680475234985, | |
| "logits/rejected": -1.9997154474258423, | |
| "logps/chosen": -29.64749526977539, | |
| "logps/rejected": -29.048025131225586, | |
| "loss": 0.5018, | |
| "rewards/accuracies": 0.42500001192092896, | |
| "rewards/chosen": -0.0026510744355618954, | |
| "rewards/margins": -0.010360640473663807, | |
| "rewards/rejected": 0.007709565572440624, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.846153846153847e-06, | |
| "logits/chosen": -1.9203827381134033, | |
| "logits/rejected": -1.9176925420761108, | |
| "logps/chosen": -31.42234230041504, | |
| "logps/rejected": -33.24127960205078, | |
| "loss": 0.4984, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.003123724367469549, | |
| "rewards/margins": 0.008287688717246056, | |
| "rewards/rejected": -0.011411413550376892, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999896948438434e-06, | |
| "logits/chosen": -2.018051862716675, | |
| "logits/rejected": -2.009334087371826, | |
| "logps/chosen": -32.55129623413086, | |
| "logps/rejected": -32.50330352783203, | |
| "loss": 0.4982, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.012754792347550392, | |
| "rewards/margins": 0.008071592077612877, | |
| "rewards/rejected": 0.004683199338614941, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.987541037542187e-06, | |
| "logits/chosen": -1.8634856939315796, | |
| "logits/rejected": -1.8527206182479858, | |
| "logps/chosen": -33.50724411010742, | |
| "logps/rejected": -35.390602111816406, | |
| "loss": 0.5016, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 0.02511655166745186, | |
| "rewards/margins": -0.005753959529101849, | |
| "rewards/rejected": 0.030870508402585983, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.954691471941119e-06, | |
| "logits/chosen": -1.9438356161117554, | |
| "logits/rejected": -1.94576096534729, | |
| "logps/chosen": -32.481632232666016, | |
| "logps/rejected": -33.15100860595703, | |
| "loss": 0.4894, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.05927763134241104, | |
| "rewards/margins": 0.049092620611190796, | |
| "rewards/rejected": 0.010185008868575096, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.901618883413549e-06, | |
| "logits/chosen": -2.0757923126220703, | |
| "logits/rejected": -2.080766439437866, | |
| "logps/chosen": -33.89708709716797, | |
| "logps/rejected": -36.524818420410156, | |
| "loss": 0.4939, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.04569761082530022, | |
| "rewards/margins": 0.023670893162488937, | |
| "rewards/rejected": 0.02202671766281128, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.828760511501322e-06, | |
| "logits/chosen": -1.936668038368225, | |
| "logits/rejected": -1.9397681951522827, | |
| "logps/chosen": -34.20936965942383, | |
| "logps/rejected": -34.525596618652344, | |
| "loss": 0.4817, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.10922203958034515, | |
| "rewards/margins": 0.07581819593906403, | |
| "rewards/rejected": 0.033403851091861725, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7367166013034295e-06, | |
| "logits/chosen": -1.9462896585464478, | |
| "logits/rejected": -1.9508006572723389, | |
| "logps/chosen": -32.27099609375, | |
| "logps/rejected": -32.275699615478516, | |
| "loss": 0.4865, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.0982138067483902, | |
| "rewards/margins": 0.05660901591181755, | |
| "rewards/rejected": 0.04160478338599205, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.626245458345211e-06, | |
| "logits/chosen": -2.043682336807251, | |
| "logits/rejected": -2.04168438911438, | |
| "logps/chosen": -31.95809555053711, | |
| "logps/rejected": -31.16133689880371, | |
| "loss": 0.4794, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.14136961102485657, | |
| "rewards/margins": 0.08893296122550964, | |
| "rewards/rejected": 0.05243664234876633, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_logits/chosen": -2.2374911308288574, | |
| "eval_logits/rejected": -2.232652425765991, | |
| "eval_logps/chosen": -33.869850158691406, | |
| "eval_logps/rejected": -37.37673568725586, | |
| "eval_loss": 0.49711015820503235, | |
| "eval_rewards/accuracies": 0.5573089718818665, | |
| "eval_rewards/chosen": 0.08235026895999908, | |
| "eval_rewards/margins": 0.012408134527504444, | |
| "eval_rewards/rejected": 0.06994213908910751, | |
| "eval_runtime": 145.9739, | |
| "eval_samples_per_second": 2.35, | |
| "eval_steps_per_second": 0.295, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.498257201263691e-06, | |
| "logits/chosen": -1.9995167255401611, | |
| "logits/rejected": -1.9971492290496826, | |
| "logps/chosen": -32.9360237121582, | |
| "logps/rejected": -33.857337951660156, | |
| "loss": 0.4819, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.15389741957187653, | |
| "rewards/margins": 0.06788130104541779, | |
| "rewards/rejected": 0.08601613342761993, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.353806263777678e-06, | |
| "logits/chosen": -2.0095458030700684, | |
| "logits/rejected": -2.001213788986206, | |
| "logps/chosen": -32.16087341308594, | |
| "logps/rejected": -31.98464584350586, | |
| "loss": 0.4876, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.142018124461174, | |
| "rewards/margins": 0.05358927324414253, | |
| "rewards/rejected": 0.08842884749174118, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.1940827077152755e-06, | |
| "logits/chosen": -2.0384786128997803, | |
| "logits/rejected": -2.030527114868164, | |
| "logps/chosen": -30.18499755859375, | |
| "logps/rejected": -31.905630111694336, | |
| "loss": 0.4841, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.14535793662071228, | |
| "rewards/margins": 0.07167014479637146, | |
| "rewards/rejected": 0.07368779182434082, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.0204024186666215e-06, | |
| "logits/chosen": -1.9675172567367554, | |
| "logits/rejected": -1.97771418094635, | |
| "logps/chosen": -31.068078994750977, | |
| "logps/rejected": -32.39047622680664, | |
| "loss": 0.4757, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.17579205334186554, | |
| "rewards/margins": 0.10358880460262299, | |
| "rewards/rejected": 0.07220325618982315, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.834196265035119e-06, | |
| "logits/chosen": -1.8799912929534912, | |
| "logits/rejected": -1.8811372518539429, | |
| "logps/chosen": -33.688819885253906, | |
| "logps/rejected": -34.5561637878418, | |
| "loss": 0.4651, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.2519363760948181, | |
| "rewards/margins": 0.15139077603816986, | |
| "rewards/rejected": 0.10054560005664825, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.636998309800573e-06, | |
| "logits/chosen": -1.9306682348251343, | |
| "logits/rejected": -1.9272987842559814, | |
| "logps/chosen": -35.7833251953125, | |
| "logps/rejected": -32.48335266113281, | |
| "loss": 0.4818, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.19417627155780792, | |
| "rewards/margins": 0.07447630167007446, | |
| "rewards/rejected": 0.11969996988773346, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4304331721118078e-06, | |
| "logits/chosen": -2.0319628715515137, | |
| "logits/rejected": -2.0246424674987793, | |
| "logps/chosen": -33.24143981933594, | |
| "logps/rejected": -31.190576553344727, | |
| "loss": 0.4536, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.2782108783721924, | |
| "rewards/margins": 0.19757375121116638, | |
| "rewards/rejected": 0.0806371346116066, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.2162026428305436e-06, | |
| "logits/chosen": -2.039504289627075, | |
| "logits/rejected": -2.044722080230713, | |
| "logps/chosen": -31.97749900817871, | |
| "logps/rejected": -32.195125579833984, | |
| "loss": 0.4662, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.28364285826683044, | |
| "rewards/margins": 0.1386357843875885, | |
| "rewards/rejected": 0.14500707387924194, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.996071664294641e-06, | |
| "logits/chosen": -2.0397489070892334, | |
| "logits/rejected": -2.037031412124634, | |
| "logps/chosen": -31.067768096923828, | |
| "logps/rejected": -31.097219467163086, | |
| "loss": 0.4756, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.21059219539165497, | |
| "rewards/margins": 0.10009355843067169, | |
| "rewards/rejected": 0.11049864441156387, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7718537898066833e-06, | |
| "logits/chosen": -1.910517930984497, | |
| "logits/rejected": -1.915186882019043, | |
| "logps/chosen": -31.059677124023438, | |
| "logps/rejected": -32.61454391479492, | |
| "loss": 0.4561, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.27330341935157776, | |
| "rewards/margins": 0.1851106435060501, | |
| "rewards/rejected": 0.08819273114204407, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_logits/chosen": -2.234231948852539, | |
| "eval_logits/rejected": -2.2294156551361084, | |
| "eval_logps/chosen": -33.776145935058594, | |
| "eval_logps/rejected": -37.29295349121094, | |
| "eval_loss": 0.49589911103248596, | |
| "eval_rewards/accuracies": 0.5423588156700134, | |
| "eval_rewards/chosen": 0.12920260429382324, | |
| "eval_rewards/margins": 0.017370687797665596, | |
| "eval_rewards/rejected": 0.1118319109082222, | |
| "eval_runtime": 145.6148, | |
| "eval_samples_per_second": 2.356, | |
| "eval_steps_per_second": 0.295, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5453962426402006e-06, | |
| "logits/chosen": -2.0229763984680176, | |
| "logits/rejected": -2.0335872173309326, | |
| "logps/chosen": -31.537296295166016, | |
| "logps/rejected": -33.72160339355469, | |
| "loss": 0.466, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.21700111031532288, | |
| "rewards/margins": 0.14645527303218842, | |
| "rewards/rejected": 0.07054580748081207, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3185646976551794e-06, | |
| "logits/chosen": -1.9160172939300537, | |
| "logits/rejected": -1.9307291507720947, | |
| "logps/chosen": -29.558719635009766, | |
| "logps/rejected": -31.404027938842773, | |
| "loss": 0.4532, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.27640971541404724, | |
| "rewards/margins": 0.19933710992336273, | |
| "rewards/rejected": 0.07707259804010391, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0932279108998323e-06, | |
| "logits/chosen": -1.9730017185211182, | |
| "logits/rejected": -1.9770148992538452, | |
| "logps/chosen": -32.84386444091797, | |
| "logps/rejected": -31.42836570739746, | |
| "loss": 0.4471, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.2868002951145172, | |
| "rewards/margins": 0.22633206844329834, | |
| "rewards/rejected": 0.06046823784708977, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8712423238279358e-06, | |
| "logits/chosen": -1.9717906713485718, | |
| "logits/rejected": -1.950059175491333, | |
| "logps/chosen": -33.563629150390625, | |
| "logps/rejected": -34.86870193481445, | |
| "loss": 0.4421, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.2950619161128998, | |
| "rewards/margins": 0.25401392579078674, | |
| "rewards/rejected": 0.041047997772693634, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6544367689701824e-06, | |
| "logits/chosen": -2.0124945640563965, | |
| "logits/rejected": -2.0092015266418457, | |
| "logps/chosen": -32.43610382080078, | |
| "logps/rejected": -35.959327697753906, | |
| "loss": 0.4732, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.24448053538799286, | |
| "rewards/margins": 0.11168196052312851, | |
| "rewards/rejected": 0.13279855251312256, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4445974030621963e-06, | |
| "logits/chosen": -1.8797550201416016, | |
| "logits/rejected": -1.8773235082626343, | |
| "logps/chosen": -33.68698501586914, | |
| "logps/rejected": -35.27508544921875, | |
| "loss": 0.4682, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.25070956349372864, | |
| "rewards/margins": 0.13594172894954681, | |
| "rewards/rejected": 0.11476783454418182, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.243452991757889e-06, | |
| "logits/chosen": -1.8655035495758057, | |
| "logits/rejected": -1.8629907369613647, | |
| "logps/chosen": -33.91318893432617, | |
| "logps/rejected": -31.576608657836914, | |
| "loss": 0.4665, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.24220648407936096, | |
| "rewards/margins": 0.1448442041873932, | |
| "rewards/rejected": 0.09736229479312897, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0526606671603523e-06, | |
| "logits/chosen": -1.9690582752227783, | |
| "logits/rejected": -1.9586395025253296, | |
| "logps/chosen": -34.75696563720703, | |
| "logps/rejected": -31.642765045166016, | |
| "loss": 0.4519, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.2998362183570862, | |
| "rewards/margins": 0.20072226226329803, | |
| "rewards/rejected": 0.09911395609378815, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.737922755071455e-07, | |
| "logits/chosen": -2.0643129348754883, | |
| "logits/rejected": -2.049489736557007, | |
| "logps/chosen": -30.391122817993164, | |
| "logps/rejected": -32.35709762573242, | |
| "loss": 0.4722, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.2665007710456848, | |
| "rewards/margins": 0.11913253366947174, | |
| "rewards/rejected": 0.14736825227737427, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.08321427484816e-07, | |
| "logits/chosen": -1.9364421367645264, | |
| "logits/rejected": -1.933985710144043, | |
| "logps/chosen": -32.0788688659668, | |
| "logps/rejected": -30.654926300048828, | |
| "loss": 0.4202, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.42412155866622925, | |
| "rewards/margins": 0.34317898750305176, | |
| "rewards/rejected": 0.08094261586666107, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_logits/chosen": -2.233308792114258, | |
| "eval_logits/rejected": -2.2285048961639404, | |
| "eval_logps/chosen": -33.75082015991211, | |
| "eval_logps/rejected": -37.27885055541992, | |
| "eval_loss": 0.4944371283054352, | |
| "eval_rewards/accuracies": 0.5307309031486511, | |
| "eval_rewards/chosen": 0.14186599850654602, | |
| "eval_rewards/margins": 0.022981125861406326, | |
| "eval_rewards/rejected": 0.1188848614692688, | |
| "eval_runtime": 145.5333, | |
| "eval_samples_per_second": 2.357, | |
| "eval_steps_per_second": 0.295, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.576113578589035e-07, | |
| "logits/chosen": -1.9192613363265991, | |
| "logits/rejected": -1.9161159992218018, | |
| "logps/chosen": -31.030216217041016, | |
| "logps/rejected": -33.573768615722656, | |
| "loss": 0.4538, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.2865663468837738, | |
| "rewards/margins": 0.20237243175506592, | |
| "rewards/rejected": 0.0841938853263855, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.229036944380913e-07, | |
| "logits/chosen": -1.9703991413116455, | |
| "logits/rejected": -1.9582570791244507, | |
| "logps/chosen": -34.00434112548828, | |
| "logps/rejected": -33.427330017089844, | |
| "loss": 0.4439, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.2958681583404541, | |
| "rewards/margins": 0.24011309444904327, | |
| "rewards/rejected": 0.05575507879257202, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.053082288996112e-07, | |
| "logits/chosen": -2.0051374435424805, | |
| "logits/rejected": -2.003786563873291, | |
| "logps/chosen": -32.876285552978516, | |
| "logps/rejected": -32.25849914550781, | |
| "loss": 0.4543, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.30475252866744995, | |
| "rewards/margins": 0.1942104697227478, | |
| "rewards/rejected": 0.11054208129644394, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.0579377374915805e-07, | |
| "logits/chosen": -2.092106580734253, | |
| "logits/rejected": -2.0764684677124023, | |
| "logps/chosen": -33.44976043701172, | |
| "logps/rejected": -32.806304931640625, | |
| "loss": 0.4513, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.3648151457309723, | |
| "rewards/margins": 0.2089495211839676, | |
| "rewards/rejected": 0.1558656245470047, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.2518018074041684e-07, | |
| "logits/chosen": -1.9645261764526367, | |
| "logits/rejected": -1.963702917098999, | |
| "logps/chosen": -32.543540954589844, | |
| "logps/rejected": -32.22251510620117, | |
| "loss": 0.4427, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.36927860975265503, | |
| "rewards/margins": 0.24796243011951447, | |
| "rewards/rejected": 0.12131617963314056, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.41315865106129e-08, | |
| "logits/chosen": -1.921724557876587, | |
| "logits/rejected": -1.9320251941680908, | |
| "logps/chosen": -31.573944091796875, | |
| "logps/rejected": -35.00251007080078, | |
| "loss": 0.4596, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.3153747618198395, | |
| "rewards/margins": 0.17225053906440735, | |
| "rewards/rejected": 0.14312422275543213, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.3150941078050325e-08, | |
| "logits/chosen": -2.059705972671509, | |
| "logits/rejected": -2.0532097816467285, | |
| "logps/chosen": -33.03301239013672, | |
| "logps/rejected": -28.99460220336914, | |
| "loss": 0.4533, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.3094797134399414, | |
| "rewards/margins": 0.19356802105903625, | |
| "rewards/rejected": 0.11591170728206635, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.575864278703266e-09, | |
| "logits/chosen": -1.919736623764038, | |
| "logits/rejected": -1.9219011068344116, | |
| "logps/chosen": -33.58899688720703, | |
| "logps/rejected": -30.707019805908203, | |
| "loss": 0.4436, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.3250266909599304, | |
| "rewards/margins": 0.24222226440906525, | |
| "rewards/rejected": 0.08280440419912338, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 385, | |
| "total_flos": 0.0, | |
| "train_loss": 0.4687722819192069, | |
| "train_runtime": 3250.758, | |
| "train_samples_per_second": 0.947, | |
| "train_steps_per_second": 0.118 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 385, | |
| "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 | |
| } | |