| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.5578800557880056, |
| "eval_steps": 300, |
| "global_step": 250, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0022315202231520223, |
| "grad_norm": 8.175573348999023, |
| "learning_rate": 4.444444444444445e-06, |
| "logits/chosen": -0.48349103331565857, |
| "logits/rejected": -0.652603030204773, |
| "logps/chosen": -8.859519004821777, |
| "logps/rejected": -16.016435623168945, |
| "loss": 1.3468515872955322, |
| "memory(GiB)": 33.44, |
| "nll_loss": 0.6537042856216431, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1, |
| "train_speed(iter/s)": 0.010093 |
| }, |
| { |
| "epoch": 0.004463040446304045, |
| "grad_norm": 6.496051788330078, |
| "learning_rate": 8.88888888888889e-06, |
| "logits/chosen": -0.5174899101257324, |
| "logits/rejected": -0.6262180805206299, |
| "logps/chosen": -9.287378311157227, |
| "logps/rejected": -13.981985092163086, |
| "loss": 1.290852665901184, |
| "memory(GiB)": 36.3, |
| "nll_loss": 0.5977055430412292, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 2, |
| "train_speed(iter/s)": 0.010436 |
| }, |
| { |
| "epoch": 0.0066945606694560665, |
| "grad_norm": 5.3426923751831055, |
| "learning_rate": 1.3333333333333333e-05, |
| "logits/chosen": -0.4254978895187378, |
| "logits/rejected": -0.6118676662445068, |
| "logps/chosen": -6.5775346755981445, |
| "logps/rejected": -16.196733474731445, |
| "loss": 1.1887776851654053, |
| "memory(GiB)": 36.3, |
| "nll_loss": 0.49580031633377075, |
| "rewards/accuracies": 0.46875, |
| "rewards/chosen": -0.0025197849608957767, |
| "rewards/margins": 0.0005687186494469643, |
| "rewards/rejected": -0.0030885031446814537, |
| "step": 3, |
| "train_speed(iter/s)": 0.010535 |
| }, |
| { |
| "epoch": 0.00892608089260809, |
| "grad_norm": 5.717321395874023, |
| "learning_rate": 1.777777777777778e-05, |
| "logits/chosen": -0.5246356129646301, |
| "logits/rejected": -0.651706337928772, |
| "logps/chosen": -7.447962760925293, |
| "logps/rejected": -16.996183395385742, |
| "loss": 1.3215211629867554, |
| "memory(GiB)": 39.38, |
| "nll_loss": 0.632453203201294, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.004622042179107666, |
| "rewards/margins": 0.008373789489269257, |
| "rewards/rejected": -0.0037517473101615906, |
| "step": 4, |
| "train_speed(iter/s)": 0.010586 |
| }, |
| { |
| "epoch": 0.011157601115760111, |
| "grad_norm": 7.188594341278076, |
| "learning_rate": 2.2222222222222223e-05, |
| "logits/chosen": -0.5062788128852844, |
| "logits/rejected": -0.6033568382263184, |
| "logps/chosen": -7.093954086303711, |
| "logps/rejected": -15.186193466186523, |
| "loss": 1.0765105485916138, |
| "memory(GiB)": 39.38, |
| "nll_loss": 0.4042052626609802, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.018988817930221558, |
| "rewards/margins": 0.04282933473587036, |
| "rewards/rejected": -0.023840520530939102, |
| "step": 5, |
| "train_speed(iter/s)": 0.010619 |
| }, |
| { |
| "epoch": 0.013389121338912133, |
| "grad_norm": 3.8232157230377197, |
| "learning_rate": 2.6666666666666667e-05, |
| "logits/chosen": -0.4826778173446655, |
| "logits/rejected": -0.5863982439041138, |
| "logps/chosen": -6.322361946105957, |
| "logps/rejected": -16.8499755859375, |
| "loss": 1.2770025730133057, |
| "memory(GiB)": 39.38, |
| "nll_loss": 0.6041065454483032, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.007987724617123604, |
| "rewards/margins": 0.0449829027056694, |
| "rewards/rejected": -0.03699517622590065, |
| "step": 6, |
| "train_speed(iter/s)": 0.010645 |
| }, |
| { |
| "epoch": 0.015620641562064157, |
| "grad_norm": 3.0141377449035645, |
| "learning_rate": 3.111111111111111e-05, |
| "logits/chosen": -0.557126522064209, |
| "logits/rejected": -0.6060188412666321, |
| "logps/chosen": -5.420467853546143, |
| "logps/rejected": -11.815977096557617, |
| "loss": 1.1219334602355957, |
| "memory(GiB)": 39.38, |
| "nll_loss": 0.46769219636917114, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.010057949461042881, |
| "rewards/margins": 0.10709290951490402, |
| "rewards/rejected": -0.09703496098518372, |
| "step": 7, |
| "train_speed(iter/s)": 0.010666 |
| }, |
| { |
| "epoch": 0.01785216178521618, |
| "grad_norm": 2.8756191730499268, |
| "learning_rate": 3.555555555555556e-05, |
| "logits/chosen": -0.5616711378097534, |
| "logits/rejected": -0.6365323066711426, |
| "logps/chosen": -6.283439636230469, |
| "logps/rejected": -19.981098175048828, |
| "loss": 0.8637306690216064, |
| "memory(GiB)": 39.38, |
| "nll_loss": 0.33207178115844727, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.011994147673249245, |
| "rewards/margins": 0.4002670347690582, |
| "rewards/rejected": -0.4122611880302429, |
| "step": 8, |
| "train_speed(iter/s)": 0.010684 |
| }, |
| { |
| "epoch": 0.0200836820083682, |
| "grad_norm": 5.291012763977051, |
| "learning_rate": 4e-05, |
| "logits/chosen": -0.502879798412323, |
| "logits/rejected": -0.5977268815040588, |
| "logps/chosen": -16.167051315307617, |
| "logps/rejected": -20.485633850097656, |
| "loss": 1.618671178817749, |
| "memory(GiB)": 39.38, |
| "nll_loss": 1.0418576002120972, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.21173161268234253, |
| "rewards/margins": 0.39757153391838074, |
| "rewards/rejected": -0.6093031167984009, |
| "step": 9, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.022315202231520222, |
| "grad_norm": 5.504859447479248, |
| "learning_rate": 4.4444444444444447e-05, |
| "logits/chosen": -0.4140257239341736, |
| "logits/rejected": -0.5436943173408508, |
| "logps/chosen": -10.317049026489258, |
| "logps/rejected": -26.558115005493164, |
| "loss": 1.271435260772705, |
| "memory(GiB)": 39.38, |
| "nll_loss": 0.6719653606414795, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.2977833151817322, |
| "rewards/margins": 0.4953876733779907, |
| "rewards/rejected": -0.7931710481643677, |
| "step": 10, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.024546722454672244, |
| "grad_norm": 8.8348970413208, |
| "learning_rate": 4.888888888888889e-05, |
| "logits/chosen": -0.4538601040840149, |
| "logits/rejected": -0.5884804129600525, |
| "logps/chosen": -15.600846290588379, |
| "logps/rejected": -27.856977462768555, |
| "loss": 1.6294305324554443, |
| "memory(GiB)": 39.38, |
| "nll_loss": 1.0609474182128906, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.48508310317993164, |
| "rewards/margins": 0.41195282340049744, |
| "rewards/rejected": -0.8970359563827515, |
| "step": 11, |
| "train_speed(iter/s)": 0.0107 |
| }, |
| { |
| "epoch": 0.026778242677824266, |
| "grad_norm": 4.871168613433838, |
| "learning_rate": 5.333333333333333e-05, |
| "logits/chosen": -0.4462035596370697, |
| "logits/rejected": -0.5622009634971619, |
| "logps/chosen": -10.827975273132324, |
| "logps/rejected": -27.8737850189209, |
| "loss": 1.309844732284546, |
| "memory(GiB)": 42.78, |
| "nll_loss": 0.6817098259925842, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": -0.34655484557151794, |
| "rewards/margins": 0.38605934381484985, |
| "rewards/rejected": -0.7326142191886902, |
| "step": 12, |
| "train_speed(iter/s)": 0.010701 |
| }, |
| { |
| "epoch": 0.02900976290097629, |
| "grad_norm": 5.154696464538574, |
| "learning_rate": 5.7777777777777776e-05, |
| "logits/chosen": -0.45840126276016235, |
| "logits/rejected": -0.5749193429946899, |
| "logps/chosen": -9.034074783325195, |
| "logps/rejected": -23.558507919311523, |
| "loss": 1.3584285974502563, |
| "memory(GiB)": 42.78, |
| "nll_loss": 0.8017209768295288, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.15672500431537628, |
| "rewards/margins": 0.44559869170188904, |
| "rewards/rejected": -0.6023237109184265, |
| "step": 13, |
| "train_speed(iter/s)": 0.010704 |
| }, |
| { |
| "epoch": 0.031241283124128313, |
| "grad_norm": 2.928802013397217, |
| "learning_rate": 6.222222222222222e-05, |
| "logits/chosen": -0.4339243173599243, |
| "logits/rejected": -0.5222603678703308, |
| "logps/chosen": -7.9224724769592285, |
| "logps/rejected": -25.263307571411133, |
| "loss": 1.176775574684143, |
| "memory(GiB)": 46.23, |
| "nll_loss": 0.548651397228241, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.09932564944028854, |
| "rewards/margins": 0.27659934759140015, |
| "rewards/rejected": -0.3759249448776245, |
| "step": 14, |
| "train_speed(iter/s)": 0.010698 |
| }, |
| { |
| "epoch": 0.03347280334728033, |
| "grad_norm": 1.9422496557235718, |
| "learning_rate": 6.666666666666667e-05, |
| "logits/chosen": -0.4753529131412506, |
| "logits/rejected": -0.5179453492164612, |
| "logps/chosen": -8.660318374633789, |
| "logps/rejected": -16.743135452270508, |
| "loss": 1.1362940073013306, |
| "memory(GiB)": 46.23, |
| "nll_loss": 0.5188332796096802, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.09274716675281525, |
| "rewards/margins": 0.20326045155525208, |
| "rewards/rejected": -0.29600760340690613, |
| "step": 15, |
| "train_speed(iter/s)": 0.010702 |
| }, |
| { |
| "epoch": 0.03570432357043236, |
| "grad_norm": 1.7832196950912476, |
| "learning_rate": 7.111111111111112e-05, |
| "logits/chosen": -0.39417293667793274, |
| "logits/rejected": -0.5316826701164246, |
| "logps/chosen": -4.724964618682861, |
| "logps/rejected": -19.69357681274414, |
| "loss": 1.0538526773452759, |
| "memory(GiB)": 46.23, |
| "nll_loss": 0.4770840108394623, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06691457331180573, |
| "rewards/margins": 0.27725517749786377, |
| "rewards/rejected": -0.21034058928489685, |
| "step": 16, |
| "train_speed(iter/s)": 0.010702 |
| }, |
| { |
| "epoch": 0.03793584379358438, |
| "grad_norm": 1.5563856363296509, |
| "learning_rate": 7.555555555555556e-05, |
| "logits/chosen": -0.4563882350921631, |
| "logits/rejected": -0.5541551113128662, |
| "logps/chosen": -9.05321979522705, |
| "logps/rejected": -18.82474708557129, |
| "loss": 1.2152374982833862, |
| "memory(GiB)": 46.23, |
| "nll_loss": 0.6201823353767395, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.11768844723701477, |
| "rewards/margins": 0.2521967887878418, |
| "rewards/rejected": -0.13450834155082703, |
| "step": 17, |
| "train_speed(iter/s)": 0.010706 |
| }, |
| { |
| "epoch": 0.0401673640167364, |
| "grad_norm": 1.9268314838409424, |
| "learning_rate": 8e-05, |
| "logits/chosen": -0.4476441740989685, |
| "logits/rejected": -0.4938126504421234, |
| "logps/chosen": -9.981496810913086, |
| "logps/rejected": -14.752214431762695, |
| "loss": 1.1654764413833618, |
| "memory(GiB)": 46.23, |
| "nll_loss": 0.46348586678504944, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.008895257487893105, |
| "rewards/margins": 0.039339594542980194, |
| "rewards/rejected": -0.030444344505667686, |
| "step": 18, |
| "train_speed(iter/s)": 0.010707 |
| }, |
| { |
| "epoch": 0.042398884239888426, |
| "grad_norm": 1.6981456279754639, |
| "learning_rate": 8.444444444444444e-05, |
| "logits/chosen": -0.49693721532821655, |
| "logits/rejected": -0.5383259057998657, |
| "logps/chosen": -7.632569313049316, |
| "logps/rejected": -17.26980209350586, |
| "loss": 1.1578692197799683, |
| "memory(GiB)": 46.23, |
| "nll_loss": 0.5097256898880005, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.06525340676307678, |
| "rewards/margins": 0.123507060110569, |
| "rewards/rejected": -0.05825363099575043, |
| "step": 19, |
| "train_speed(iter/s)": 0.010709 |
| }, |
| { |
| "epoch": 0.044630404463040445, |
| "grad_norm": 2.013322591781616, |
| "learning_rate": 8.888888888888889e-05, |
| "logits/chosen": -0.4885023236274719, |
| "logits/rejected": -0.5652343034744263, |
| "logps/chosen": -6.508520603179932, |
| "logps/rejected": -13.635923385620117, |
| "loss": 1.0892918109893799, |
| "memory(GiB)": 46.23, |
| "nll_loss": 0.45972177386283875, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.06720655411481857, |
| "rewards/margins": 0.18712735176086426, |
| "rewards/rejected": -0.11992079019546509, |
| "step": 20, |
| "train_speed(iter/s)": 0.010713 |
| }, |
| { |
| "epoch": 0.04686192468619247, |
| "grad_norm": 1.775608777999878, |
| "learning_rate": 9.333333333333334e-05, |
| "logits/chosen": -0.4235825836658478, |
| "logits/rejected": -0.4997040629386902, |
| "logps/chosen": -7.756269454956055, |
| "logps/rejected": -22.05597496032715, |
| "loss": 1.1610935926437378, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5124741792678833, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.056686438620090485, |
| "rewards/margins": 0.15342777967453003, |
| "rewards/rejected": -0.09674134850502014, |
| "step": 21, |
| "train_speed(iter/s)": 0.010711 |
| }, |
| { |
| "epoch": 0.04909344490934449, |
| "grad_norm": 1.6878494024276733, |
| "learning_rate": 9.777777777777778e-05, |
| "logits/chosen": -0.448173850774765, |
| "logits/rejected": -0.46942824125289917, |
| "logps/chosen": -9.774542808532715, |
| "logps/rejected": -14.852673530578613, |
| "loss": 1.1857895851135254, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5429187417030334, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.08716461062431335, |
| "rewards/margins": 0.17264463007450104, |
| "rewards/rejected": -0.08548003435134888, |
| "step": 22, |
| "train_speed(iter/s)": 0.01071 |
| }, |
| { |
| "epoch": 0.051324965132496514, |
| "grad_norm": 1.9283621311187744, |
| "learning_rate": 0.00010222222222222222, |
| "logits/chosen": -0.4297652244567871, |
| "logits/rejected": -0.5464367866516113, |
| "logps/chosen": -7.680075168609619, |
| "logps/rejected": -20.64339828491211, |
| "loss": 1.1528663635253906, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.48048245906829834, |
| "rewards/accuracies": 0.46875, |
| "rewards/chosen": 0.05369298905134201, |
| "rewards/margins": 0.09629381448030472, |
| "rewards/rejected": -0.04260082542896271, |
| "step": 23, |
| "train_speed(iter/s)": 0.01071 |
| }, |
| { |
| "epoch": 0.05355648535564853, |
| "grad_norm": 1.8897244930267334, |
| "learning_rate": 0.00010666666666666667, |
| "logits/chosen": -0.4073754847049713, |
| "logits/rejected": -0.4649578034877777, |
| "logps/chosen": -9.451826095581055, |
| "logps/rejected": -20.540328979492188, |
| "loss": 1.0510568618774414, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.431056946516037, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.055333852767944336, |
| "rewards/margins": 0.23058435320854187, |
| "rewards/rejected": -0.17525050044059753, |
| "step": 24, |
| "train_speed(iter/s)": 0.010709 |
| }, |
| { |
| "epoch": 0.05578800557880056, |
| "grad_norm": 1.968520998954773, |
| "learning_rate": 0.00011111111111111112, |
| "logits/chosen": -0.4085446000099182, |
| "logits/rejected": -0.49927735328674316, |
| "logps/chosen": -9.85863971710205, |
| "logps/rejected": -21.732637405395508, |
| "loss": 1.166273593902588, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5423365831375122, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.03325103968381882, |
| "rewards/margins": 0.2285088747739792, |
| "rewards/rejected": -0.19525781273841858, |
| "step": 25, |
| "train_speed(iter/s)": 0.010709 |
| }, |
| { |
| "epoch": 0.05801952580195258, |
| "grad_norm": 3.0135302543640137, |
| "learning_rate": 0.00011555555555555555, |
| "logits/chosen": -0.44835054874420166, |
| "logits/rejected": -0.5419439077377319, |
| "logps/chosen": -4.6747965812683105, |
| "logps/rejected": -18.435997009277344, |
| "loss": 0.9960591793060303, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.39432889223098755, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.017948223277926445, |
| "rewards/margins": 0.27569395303726196, |
| "rewards/rejected": -0.29364219307899475, |
| "step": 26, |
| "train_speed(iter/s)": 0.010712 |
| }, |
| { |
| "epoch": 0.0602510460251046, |
| "grad_norm": 2.0994958877563477, |
| "learning_rate": 0.00012, |
| "logits/chosen": -0.4755466878414154, |
| "logits/rejected": -0.513874888420105, |
| "logps/chosen": -14.17818832397461, |
| "logps/rejected": -15.192220687866211, |
| "loss": 1.3128914833068848, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.732846736907959, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.11946821957826614, |
| "rewards/margins": 0.27503034472465515, |
| "rewards/rejected": -0.1555621325969696, |
| "step": 27, |
| "train_speed(iter/s)": 0.010714 |
| }, |
| { |
| "epoch": 0.06248256624825663, |
| "grad_norm": 2.619943857192993, |
| "learning_rate": 0.00012444444444444444, |
| "logits/chosen": -0.33525484800338745, |
| "logits/rejected": -0.4649803340435028, |
| "logps/chosen": -6.553568363189697, |
| "logps/rejected": -23.942062377929688, |
| "loss": 1.007144808769226, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4446461498737335, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.07584571838378906, |
| "rewards/margins": 0.3631606996059418, |
| "rewards/rejected": -0.2873149812221527, |
| "step": 28, |
| "train_speed(iter/s)": 0.010711 |
| }, |
| { |
| "epoch": 0.06471408647140865, |
| "grad_norm": 2.1066770553588867, |
| "learning_rate": 0.00012888888888888892, |
| "logits/chosen": -0.37913525104522705, |
| "logits/rejected": -0.4620123505592346, |
| "logps/chosen": -7.321231842041016, |
| "logps/rejected": -21.526382446289062, |
| "loss": 1.0192487239837646, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4804428219795227, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.026389580219984055, |
| "rewards/margins": 0.46746930480003357, |
| "rewards/rejected": -0.441079705953598, |
| "step": 29, |
| "train_speed(iter/s)": 0.010712 |
| }, |
| { |
| "epoch": 0.06694560669456066, |
| "grad_norm": 1.8226919174194336, |
| "learning_rate": 0.00013333333333333334, |
| "logits/chosen": -0.3827905058860779, |
| "logits/rejected": -0.48159855604171753, |
| "logps/chosen": -6.776709079742432, |
| "logps/rejected": -22.80392074584961, |
| "loss": 1.0347448587417603, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.500119149684906, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.10474913567304611, |
| "rewards/margins": 0.39701035618782043, |
| "rewards/rejected": -0.2922612428665161, |
| "step": 30, |
| "train_speed(iter/s)": 0.010712 |
| }, |
| { |
| "epoch": 0.06917712691771269, |
| "grad_norm": 1.8681020736694336, |
| "learning_rate": 0.0001377777777777778, |
| "logits/chosen": -0.3609718978404999, |
| "logits/rejected": -0.43842607736587524, |
| "logps/chosen": -5.881144046783447, |
| "logps/rejected": -21.236047744750977, |
| "loss": 0.9905272126197815, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.43477141857147217, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.04010690003633499, |
| "rewards/margins": 0.3823702037334442, |
| "rewards/rejected": -0.34226328134536743, |
| "step": 31, |
| "train_speed(iter/s)": 0.010714 |
| }, |
| { |
| "epoch": 0.07140864714086471, |
| "grad_norm": 2.225977897644043, |
| "learning_rate": 0.00014222222222222224, |
| "logits/chosen": -0.39412301778793335, |
| "logits/rejected": -0.4322904050350189, |
| "logps/chosen": -10.140836715698242, |
| "logps/rejected": -18.79208755493164, |
| "loss": 1.0995820760726929, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5035545229911804, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.04120786115527153, |
| "rewards/margins": 0.2624339461326599, |
| "rewards/rejected": -0.22122608125209808, |
| "step": 32, |
| "train_speed(iter/s)": 0.010716 |
| }, |
| { |
| "epoch": 0.07364016736401674, |
| "grad_norm": 2.788853406906128, |
| "learning_rate": 0.00014666666666666666, |
| "logits/chosen": -0.27406275272369385, |
| "logits/rejected": -0.371414452791214, |
| "logps/chosen": -12.10950756072998, |
| "logps/rejected": -24.930038452148438, |
| "loss": 1.2428903579711914, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6102337837219238, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.08664405345916748, |
| "rewards/margins": 0.2861720621585846, |
| "rewards/rejected": -0.3728161156177521, |
| "step": 33, |
| "train_speed(iter/s)": 0.010715 |
| }, |
| { |
| "epoch": 0.07587168758716877, |
| "grad_norm": 2.416369676589966, |
| "learning_rate": 0.0001511111111111111, |
| "logits/chosen": -0.24224025011062622, |
| "logits/rejected": -0.3833686113357544, |
| "logps/chosen": -6.028831958770752, |
| "logps/rejected": -22.455066680908203, |
| "loss": 0.9914643168449402, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.44179052114486694, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.011510152369737625, |
| "rewards/margins": 0.5171858668327332, |
| "rewards/rejected": -0.5056756734848022, |
| "step": 34, |
| "train_speed(iter/s)": 0.010714 |
| }, |
| { |
| "epoch": 0.07810320781032078, |
| "grad_norm": 4.029798984527588, |
| "learning_rate": 0.00015555555555555556, |
| "logits/chosen": -0.2615310847759247, |
| "logits/rejected": -0.39800190925598145, |
| "logps/chosen": -5.90588903427124, |
| "logps/rejected": -24.642311096191406, |
| "loss": 0.8372505307197571, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4036243259906769, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.03906700387597084, |
| "rewards/margins": 0.8896325826644897, |
| "rewards/rejected": -0.9286996126174927, |
| "step": 35, |
| "train_speed(iter/s)": 0.010715 |
| }, |
| { |
| "epoch": 0.0803347280334728, |
| "grad_norm": 2.4356110095977783, |
| "learning_rate": 0.00016, |
| "logits/chosen": -0.25296953320503235, |
| "logits/rejected": -0.3325399160385132, |
| "logps/chosen": -9.764960289001465, |
| "logps/rejected": -23.14637565612793, |
| "loss": 1.0952283143997192, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.641168475151062, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.002499673515558243, |
| "rewards/margins": 0.8523399829864502, |
| "rewards/rejected": -0.8498403429985046, |
| "step": 36, |
| "train_speed(iter/s)": 0.010716 |
| }, |
| { |
| "epoch": 0.08256624825662483, |
| "grad_norm": 1.9291634559631348, |
| "learning_rate": 0.00016444444444444444, |
| "logits/chosen": -0.20930960774421692, |
| "logits/rejected": -0.3604508936405182, |
| "logps/chosen": -2.847867488861084, |
| "logps/rejected": -30.63514518737793, |
| "loss": 0.6885566711425781, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.2891227900981903, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.061974309384822845, |
| "rewards/margins": 1.0785088539123535, |
| "rewards/rejected": -1.0165345668792725, |
| "step": 37, |
| "train_speed(iter/s)": 0.010715 |
| }, |
| { |
| "epoch": 0.08479776847977685, |
| "grad_norm": 1.6937510967254639, |
| "learning_rate": 0.00016888888888888889, |
| "logits/chosen": -0.2987847626209259, |
| "logits/rejected": -0.37470635771751404, |
| "logps/chosen": -6.199091911315918, |
| "logps/rejected": -19.3011474609375, |
| "loss": 0.8923996090888977, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.38914668560028076, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0021905135363340378, |
| "rewards/margins": 0.6954408288002014, |
| "rewards/rejected": -0.6932503581047058, |
| "step": 38, |
| "train_speed(iter/s)": 0.010716 |
| }, |
| { |
| "epoch": 0.08702928870292886, |
| "grad_norm": 2.106278896331787, |
| "learning_rate": 0.00017333333333333334, |
| "logits/chosen": -0.28487277030944824, |
| "logits/rejected": -0.34711676836013794, |
| "logps/chosen": -8.657167434692383, |
| "logps/rejected": -26.564130783081055, |
| "loss": 1.023146390914917, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5919384956359863, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.05617165565490723, |
| "rewards/margins": 1.1481283903121948, |
| "rewards/rejected": -1.091956615447998, |
| "step": 39, |
| "train_speed(iter/s)": 0.010717 |
| }, |
| { |
| "epoch": 0.08926080892608089, |
| "grad_norm": 2.619154214859009, |
| "learning_rate": 0.00017777777777777779, |
| "logits/chosen": -0.2898882031440735, |
| "logits/rejected": -0.3393169343471527, |
| "logps/chosen": -8.001999855041504, |
| "logps/rejected": -20.189369201660156, |
| "loss": 0.9479650855064392, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.42854148149490356, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.0008942075073719025, |
| "rewards/margins": 0.7672775983810425, |
| "rewards/rejected": -0.7663832902908325, |
| "step": 40, |
| "train_speed(iter/s)": 0.010719 |
| }, |
| { |
| "epoch": 0.09149232914923291, |
| "grad_norm": 2.770841360092163, |
| "learning_rate": 0.00018222222222222224, |
| "logits/chosen": -0.22048279643058777, |
| "logits/rejected": -0.30986371636390686, |
| "logps/chosen": -5.240083694458008, |
| "logps/rejected": -22.307451248168945, |
| "loss": 0.8702710866928101, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.39226293563842773, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.017199480906128883, |
| "rewards/margins": 0.9372033476829529, |
| "rewards/rejected": -0.9200038313865662, |
| "step": 41, |
| "train_speed(iter/s)": 0.010719 |
| }, |
| { |
| "epoch": 0.09372384937238494, |
| "grad_norm": 4.61620569229126, |
| "learning_rate": 0.0001866666666666667, |
| "logits/chosen": -0.30797824263572693, |
| "logits/rejected": -0.3721832036972046, |
| "logps/chosen": -10.481635093688965, |
| "logps/rejected": -23.132564544677734, |
| "loss": 1.2510943412780762, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6913026571273804, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.25621044635772705, |
| "rewards/margins": 0.5585607886314392, |
| "rewards/rejected": -0.814771294593811, |
| "step": 42, |
| "train_speed(iter/s)": 0.010721 |
| }, |
| { |
| "epoch": 0.09595536959553697, |
| "grad_norm": 3.806872844696045, |
| "learning_rate": 0.00019111111111111114, |
| "logits/chosen": -0.22567886114120483, |
| "logits/rejected": -0.3761019706726074, |
| "logps/chosen": -8.201981544494629, |
| "logps/rejected": -34.75733184814453, |
| "loss": 1.0068457126617432, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6396302580833435, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.12671321630477905, |
| "rewards/margins": 1.7875899076461792, |
| "rewards/rejected": -1.914302945137024, |
| "step": 43, |
| "train_speed(iter/s)": 0.010722 |
| }, |
| { |
| "epoch": 0.09818688981868898, |
| "grad_norm": 3.342564821243286, |
| "learning_rate": 0.00019555555555555556, |
| "logits/chosen": -0.259334534406662, |
| "logits/rejected": -0.33293721079826355, |
| "logps/chosen": -9.735747337341309, |
| "logps/rejected": -34.48583221435547, |
| "loss": 0.9845027327537537, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5914844870567322, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.2998116612434387, |
| "rewards/margins": 1.7843880653381348, |
| "rewards/rejected": -2.084199905395508, |
| "step": 44, |
| "train_speed(iter/s)": 0.010723 |
| }, |
| { |
| "epoch": 0.100418410041841, |
| "grad_norm": 3.4293951988220215, |
| "learning_rate": 0.0002, |
| "logits/chosen": -0.34163057804107666, |
| "logits/rejected": -0.36223551630973816, |
| "logps/chosen": -9.353399276733398, |
| "logps/rejected": -21.651199340820312, |
| "loss": 1.0366103649139404, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5141280889511108, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.003630978986620903, |
| "rewards/margins": 0.7049190402030945, |
| "rewards/rejected": -0.7085499167442322, |
| "step": 45, |
| "train_speed(iter/s)": 0.010724 |
| }, |
| { |
| "epoch": 0.10264993026499303, |
| "grad_norm": 2.1774137020111084, |
| "learning_rate": 0.0001999993185874369, |
| "logits/chosen": -0.3677416443824768, |
| "logits/rejected": -0.3854161500930786, |
| "logps/chosen": -11.536070823669434, |
| "logps/rejected": -27.305158615112305, |
| "loss": 0.9515228271484375, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5171921849250793, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.09924840927124023, |
| "rewards/margins": 1.116947889328003, |
| "rewards/rejected": -1.2161962985992432, |
| "step": 46, |
| "train_speed(iter/s)": 0.010724 |
| }, |
| { |
| "epoch": 0.10488145048814505, |
| "grad_norm": 2.2629730701446533, |
| "learning_rate": 0.00019999727435903407, |
| "logits/chosen": -0.3420083522796631, |
| "logits/rejected": -0.4473438858985901, |
| "logps/chosen": -4.170689582824707, |
| "logps/rejected": -21.129119873046875, |
| "loss": 0.9114863276481628, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3966943621635437, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1746423840522766, |
| "rewards/margins": 0.61202472448349, |
| "rewards/rejected": -0.43738237023353577, |
| "step": 47, |
| "train_speed(iter/s)": 0.010725 |
| }, |
| { |
| "epoch": 0.10711297071129706, |
| "grad_norm": 1.6574047803878784, |
| "learning_rate": 0.0001999938673426507, |
| "logits/chosen": -0.31779757142066956, |
| "logits/rejected": -0.4338604807853699, |
| "logps/chosen": -5.825237274169922, |
| "logps/rejected": -20.981552124023438, |
| "loss": 0.9128789305686951, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4218422770500183, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1773630529642105, |
| "rewards/margins": 0.718612551689148, |
| "rewards/rejected": -0.5412494540214539, |
| "step": 48, |
| "train_speed(iter/s)": 0.010725 |
| }, |
| { |
| "epoch": 0.10934449093444909, |
| "grad_norm": 2.0683908462524414, |
| "learning_rate": 0.00019998909758471852, |
| "logits/chosen": -0.3863949179649353, |
| "logits/rejected": -0.4458945393562317, |
| "logps/chosen": -7.798675537109375, |
| "logps/rejected": -21.307409286499023, |
| "loss": 1.0268117189407349, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5051021575927734, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.08097527176141739, |
| "rewards/margins": 0.7529874444007874, |
| "rewards/rejected": -0.672012209892273, |
| "step": 49, |
| "train_speed(iter/s)": 0.010726 |
| }, |
| { |
| "epoch": 0.11157601115760112, |
| "grad_norm": 2.338461399078369, |
| "learning_rate": 0.000199982965150241, |
| "logits/chosen": -0.34079253673553467, |
| "logits/rejected": -0.3965541124343872, |
| "logps/chosen": -11.766353607177734, |
| "logps/rejected": -19.651931762695312, |
| "loss": 1.1824036836624146, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.609149158000946, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.08129046857357025, |
| "rewards/margins": 0.3777032196521759, |
| "rewards/rejected": -0.4589936435222626, |
| "step": 50, |
| "train_speed(iter/s)": 0.010726 |
| }, |
| { |
| "epoch": 0.11380753138075314, |
| "grad_norm": 1.6744564771652222, |
| "learning_rate": 0.00019997547012279244, |
| "logits/chosen": -0.31765079498291016, |
| "logits/rejected": -0.42205262184143066, |
| "logps/chosen": -5.897686958312988, |
| "logps/rejected": -25.42192268371582, |
| "loss": 0.8473464846611023, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4084213376045227, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2017829716205597, |
| "rewards/margins": 0.7832211256027222, |
| "rewards/rejected": -0.5814382433891296, |
| "step": 51, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.11603905160390517, |
| "grad_norm": 1.8061268329620361, |
| "learning_rate": 0.00019996661260451704, |
| "logits/chosen": -0.32708457112312317, |
| "logits/rejected": -0.41108813881874084, |
| "logps/chosen": -8.663649559020996, |
| "logps/rejected": -32.308448791503906, |
| "loss": 0.8938776850700378, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.48662397265434265, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.01828554831445217, |
| "rewards/margins": 1.3651036024093628, |
| "rewards/rejected": -1.3833892345428467, |
| "step": 52, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.11827057182705718, |
| "grad_norm": 1.9129594564437866, |
| "learning_rate": 0.00019995639271612722, |
| "logits/chosen": -0.3344545364379883, |
| "logits/rejected": -0.43980082869529724, |
| "logps/chosen": -9.373285293579102, |
| "logps/rejected": -30.374614715576172, |
| "loss": 0.8187388181686401, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3988412916660309, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.04744601994752884, |
| "rewards/margins": 1.1539989709854126, |
| "rewards/rejected": -1.2014449834823608, |
| "step": 53, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.1205020920502092, |
| "grad_norm": 6.22075891494751, |
| "learning_rate": 0.00019994481059690223, |
| "logits/chosen": -0.3710540533065796, |
| "logits/rejected": -0.45295441150665283, |
| "logps/chosen": -11.353669166564941, |
| "logps/rejected": -35.60136795043945, |
| "loss": 1.3072545528411865, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.8566557168960571, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.1803397834300995, |
| "rewards/margins": 2.1112091541290283, |
| "rewards/rejected": -2.29154896736145, |
| "step": 54, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.12273361227336123, |
| "grad_norm": 3.538923501968384, |
| "learning_rate": 0.00019993186640468604, |
| "logits/chosen": -0.2791898250579834, |
| "logits/rejected": -0.3996254801750183, |
| "logps/chosen": -12.779634475708008, |
| "logps/rejected": -37.21310043334961, |
| "loss": 1.0243951082229614, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6127378940582275, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.261193186044693, |
| "rewards/margins": 1.434675693511963, |
| "rewards/rejected": -1.695868968963623, |
| "step": 55, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.12496513249651325, |
| "grad_norm": 2.5944247245788574, |
| "learning_rate": 0.0001999175603158854, |
| "logits/chosen": -0.2531765401363373, |
| "logits/rejected": -0.4114147424697876, |
| "logps/chosen": -9.740252494812012, |
| "logps/rejected": -36.96112823486328, |
| "loss": 0.8653287291526794, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5133413076400757, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.015189380384981632, |
| "rewards/margins": 1.5309333801269531, |
| "rewards/rejected": -1.5461229085922241, |
| "step": 56, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.12719665271966527, |
| "grad_norm": 3.273171901702881, |
| "learning_rate": 0.0001999018925254673, |
| "logits/chosen": -0.3509857654571533, |
| "logits/rejected": -0.4661892056465149, |
| "logps/chosen": -8.814093589782715, |
| "logps/rejected": -30.390697479248047, |
| "loss": 0.9130730032920837, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5733908414840698, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.0046962546184659, |
| "rewards/margins": 1.3072673082351685, |
| "rewards/rejected": -1.3025707006454468, |
| "step": 57, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.1294281729428173, |
| "grad_norm": 2.717229127883911, |
| "learning_rate": 0.00019988486324695628, |
| "logits/chosen": -0.33554819226264954, |
| "logits/rejected": -0.4281303584575653, |
| "logps/chosen": -14.103944778442383, |
| "logps/rejected": -32.3876838684082, |
| "loss": 0.9715602397918701, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5424861311912537, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.0524468794465065, |
| "rewards/margins": 1.0883628129959106, |
| "rewards/rejected": -1.1408096551895142, |
| "step": 58, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.13165969316596932, |
| "grad_norm": 2.887491226196289, |
| "learning_rate": 0.00019986647271243163, |
| "logits/chosen": -0.3923209011554718, |
| "logits/rejected": -0.48743686079978943, |
| "logps/chosen": -6.2308878898620605, |
| "logps/rejected": -21.537296295166016, |
| "loss": 0.9816198945045471, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4462883472442627, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.06832532584667206, |
| "rewards/margins": 0.5661296844482422, |
| "rewards/rejected": -0.6344550251960754, |
| "step": 59, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.13389121338912133, |
| "grad_norm": 2.2311718463897705, |
| "learning_rate": 0.00019984672117252423, |
| "logits/chosen": -0.33507245779037476, |
| "logits/rejected": -0.49380406737327576, |
| "logps/chosen": -9.220368385314941, |
| "logps/rejected": -37.96326446533203, |
| "loss": 0.9401113986968994, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5612199902534485, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.06330323219299316, |
| "rewards/margins": 1.686211347579956, |
| "rewards/rejected": -1.622908115386963, |
| "step": 60, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.13612273361227337, |
| "grad_norm": 1.992959976196289, |
| "learning_rate": 0.00019982560889641296, |
| "logits/chosen": -0.35564690828323364, |
| "logits/rejected": -0.5200420618057251, |
| "logps/chosen": -3.4553239345550537, |
| "logps/rejected": -32.64472198486328, |
| "loss": 0.640341579914093, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3033585250377655, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.08748181164264679, |
| "rewards/margins": 1.5070304870605469, |
| "rewards/rejected": -1.419548511505127, |
| "step": 61, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.13835425383542538, |
| "grad_norm": 3.40260648727417, |
| "learning_rate": 0.00019980313617182127, |
| "logits/chosen": -0.3457183837890625, |
| "logits/rejected": -0.5498067736625671, |
| "logps/chosen": -6.69366455078125, |
| "logps/rejected": -37.40818405151367, |
| "loss": 0.9436236619949341, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6088553667068481, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.10982945561408997, |
| "rewards/margins": 1.797737956047058, |
| "rewards/rejected": -1.9075673818588257, |
| "step": 62, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.14058577405857742, |
| "grad_norm": 4.374825954437256, |
| "learning_rate": 0.00019977930330501308, |
| "logits/chosen": -0.42058196663856506, |
| "logits/rejected": -0.5126227736473083, |
| "logps/chosen": -10.140219688415527, |
| "logps/rejected": -33.032615661621094, |
| "loss": 1.193029522895813, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6540992856025696, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.22300566732883453, |
| "rewards/margins": 1.272392988204956, |
| "rewards/rejected": -1.495398759841919, |
| "step": 63, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.14281729428172943, |
| "grad_norm": 4.756910800933838, |
| "learning_rate": 0.0001997541106207887, |
| "logits/chosen": -0.4631284475326538, |
| "logits/rejected": -0.5332682728767395, |
| "logps/chosen": -8.054980278015137, |
| "logps/rejected": -31.029682159423828, |
| "loss": 1.0884121656417847, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6862107515335083, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.1161859929561615, |
| "rewards/margins": 1.4922480583190918, |
| "rewards/rejected": -1.6084339618682861, |
| "step": 64, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.14504881450488144, |
| "grad_norm": 3.428164005279541, |
| "learning_rate": 0.0001997275584624803, |
| "logits/chosen": -0.3415633738040924, |
| "logits/rejected": -0.49688977003097534, |
| "logps/chosen": -7.313436031341553, |
| "logps/rejected": -35.58869934082031, |
| "loss": 0.957207977771759, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.590711772441864, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.05026661604642868, |
| "rewards/margins": 1.5162088871002197, |
| "rewards/rejected": -1.4659425020217896, |
| "step": 65, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.14728033472803348, |
| "grad_norm": 3.1654562950134277, |
| "learning_rate": 0.00019969964719194745, |
| "logits/chosen": -0.4512020945549011, |
| "logits/rejected": -0.5287590026855469, |
| "logps/chosen": -8.630544662475586, |
| "logps/rejected": -15.341949462890625, |
| "loss": 1.1465851068496704, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5758541822433472, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.06041776388883591, |
| "rewards/margins": 0.51249760389328, |
| "rewards/rejected": -0.45207977294921875, |
| "step": 66, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.1495118549511855, |
| "grad_norm": 2.0986812114715576, |
| "learning_rate": 0.0001996703771895719, |
| "logits/chosen": -0.5272831320762634, |
| "logits/rejected": -0.517794668674469, |
| "logps/chosen": -12.980257034301758, |
| "logps/rejected": -18.752052307128906, |
| "loss": 1.3610563278198242, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6797517538070679, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.026422299444675446, |
| "rewards/margins": 0.5318213105201721, |
| "rewards/rejected": -0.5053990483283997, |
| "step": 67, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.15174337517433753, |
| "grad_norm": 1.9081969261169434, |
| "learning_rate": 0.00019963974885425266, |
| "logits/chosen": -0.4080069661140442, |
| "logits/rejected": -0.47813406586647034, |
| "logps/chosen": -7.824826717376709, |
| "logps/rejected": -18.751680374145508, |
| "loss": 1.0740342140197754, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.45793265104293823, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.04187186062335968, |
| "rewards/margins": 0.3901542127132416, |
| "rewards/rejected": -0.3482823669910431, |
| "step": 68, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.15397489539748954, |
| "grad_norm": 1.700203537940979, |
| "learning_rate": 0.0001996077626034003, |
| "logits/chosen": -0.45641785860061646, |
| "logits/rejected": -0.5060834884643555, |
| "logps/chosen": -11.462532043457031, |
| "logps/rejected": -16.27201271057129, |
| "loss": 1.1681339740753174, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6015636920928955, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.136613130569458, |
| "rewards/margins": 0.3738591969013214, |
| "rewards/rejected": -0.2372460514307022, |
| "step": 69, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.15620641562064155, |
| "grad_norm": 1.9888875484466553, |
| "learning_rate": 0.00019957441887293156, |
| "logits/chosen": -0.4031111001968384, |
| "logits/rejected": -0.47320762276649475, |
| "logps/chosen": -7.325850009918213, |
| "logps/rejected": -25.553611755371094, |
| "loss": 1.002021074295044, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4237433671951294, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.013660086318850517, |
| "rewards/margins": 0.49219831824302673, |
| "rewards/rejected": -0.4785382151603699, |
| "step": 70, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.1584379358437936, |
| "grad_norm": 1.5379258394241333, |
| "learning_rate": 0.0001995397181172631, |
| "logits/chosen": -0.40455764532089233, |
| "logits/rejected": -0.518138587474823, |
| "logps/chosen": -3.8365478515625, |
| "logps/rejected": -26.37249755859375, |
| "loss": 0.8130152225494385, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.34526968002319336, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.10106083005666733, |
| "rewards/margins": 0.8268870115280151, |
| "rewards/rejected": -0.7258262038230896, |
| "step": 71, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.1606694560669456, |
| "grad_norm": 1.612131953239441, |
| "learning_rate": 0.0001995036608093056, |
| "logits/chosen": -0.4558025002479553, |
| "logits/rejected": -0.48799094557762146, |
| "logps/chosen": -8.70871639251709, |
| "logps/rejected": -17.577770233154297, |
| "loss": 0.8906623125076294, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3765795826911926, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.05890116095542908, |
| "rewards/margins": 0.6036397814750671, |
| "rewards/rejected": -0.5447385907173157, |
| "step": 72, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.16290097629009762, |
| "grad_norm": 1.3791937828063965, |
| "learning_rate": 0.00019946624744045704, |
| "logits/chosen": -0.4362650215625763, |
| "logits/rejected": -0.48206600546836853, |
| "logps/chosen": -10.413619995117188, |
| "logps/rejected": -27.357479095458984, |
| "loss": 0.8979863524436951, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3758661448955536, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.009980186820030212, |
| "rewards/margins": 0.9504665732383728, |
| "rewards/rejected": -0.960446834564209, |
| "step": 73, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.16513249651324965, |
| "grad_norm": 1.7471097707748413, |
| "learning_rate": 0.00019942747852059632, |
| "logits/chosen": -0.423688143491745, |
| "logits/rejected": -0.49418017268180847, |
| "logps/chosen": -8.749202728271484, |
| "logps/rejected": -33.218692779541016, |
| "loss": 0.8359287977218628, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.36860865354537964, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.05798308178782463, |
| "rewards/margins": 1.4300649166107178, |
| "rewards/rejected": -1.4880479574203491, |
| "step": 74, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.16736401673640167, |
| "grad_norm": 2.2621185779571533, |
| "learning_rate": 0.00019938735457807592, |
| "logits/chosen": -0.40228813886642456, |
| "logits/rejected": -0.47656214237213135, |
| "logps/chosen": -10.918057441711426, |
| "logps/rejected": -30.61962890625, |
| "loss": 0.9725239872932434, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.499444842338562, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.06699477136135101, |
| "rewards/margins": 0.938097357749939, |
| "rewards/rejected": -1.0050921440124512, |
| "step": 75, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.1695955369595537, |
| "grad_norm": 2.474619150161743, |
| "learning_rate": 0.00019934587615971506, |
| "logits/chosen": -0.38186344504356384, |
| "logits/rejected": -0.4598054587841034, |
| "logps/chosen": -6.823958873748779, |
| "logps/rejected": -29.09862518310547, |
| "loss": 0.9761800765991211, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.48571133613586426, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.05072644352912903, |
| "rewards/margins": 0.8625801205635071, |
| "rewards/rejected": -0.9133066534996033, |
| "step": 76, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.17182705718270572, |
| "grad_norm": 1.823959231376648, |
| "learning_rate": 0.00019930304383079204, |
| "logits/chosen": -0.4149280786514282, |
| "logits/rejected": -0.5062562227249146, |
| "logps/chosen": -8.54432201385498, |
| "logps/rejected": -29.242015838623047, |
| "loss": 0.8722183108329773, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4606979191303253, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.01995794102549553, |
| "rewards/margins": 1.4311785697937012, |
| "rewards/rejected": -1.4112205505371094, |
| "step": 77, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.17405857740585773, |
| "grad_norm": 1.793901801109314, |
| "learning_rate": 0.00019925885817503663, |
| "logits/chosen": -0.4355413317680359, |
| "logits/rejected": -0.5245873332023621, |
| "logps/chosen": -4.893879413604736, |
| "logps/rejected": -25.42880630493164, |
| "loss": 0.7558175921440125, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.38899344205856323, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.08300305157899857, |
| "rewards/margins": 1.0970323085784912, |
| "rewards/rejected": -1.0140292644500732, |
| "step": 78, |
| "train_speed(iter/s)": 0.010696 |
| }, |
| { |
| "epoch": 0.17629009762900977, |
| "grad_norm": 2.114241123199463, |
| "learning_rate": 0.00019921331979462197, |
| "logits/chosen": -0.3358357548713684, |
| "logits/rejected": -0.47999072074890137, |
| "logps/chosen": -5.950570583343506, |
| "logps/rejected": -28.20298957824707, |
| "loss": 0.8057999610900879, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4144690930843353, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.05946407839655876, |
| "rewards/margins": 1.0457231998443604, |
| "rewards/rejected": -0.9862591028213501, |
| "step": 79, |
| "train_speed(iter/s)": 0.010697 |
| }, |
| { |
| "epoch": 0.17852161785216178, |
| "grad_norm": 2.7315855026245117, |
| "learning_rate": 0.00019916642931015662, |
| "logits/chosen": -0.43205446004867554, |
| "logits/rejected": -0.5006397366523743, |
| "logps/chosen": -7.170259952545166, |
| "logps/rejected": -22.804529190063477, |
| "loss": 1.0863537788391113, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6158946752548218, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.03549172729253769, |
| "rewards/margins": 0.8894610404968262, |
| "rewards/rejected": -0.9249527454376221, |
| "step": 80, |
| "train_speed(iter/s)": 0.010698 |
| }, |
| { |
| "epoch": 0.18075313807531382, |
| "grad_norm": 2.479121685028076, |
| "learning_rate": 0.00019911818736067586, |
| "logits/chosen": -0.40614452958106995, |
| "logits/rejected": -0.5118440389633179, |
| "logps/chosen": -4.780482769012451, |
| "logps/rejected": -22.085041046142578, |
| "loss": 0.7858155965805054, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.44690829515457153, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.21181246638298035, |
| "rewards/margins": 1.2262333631515503, |
| "rewards/rejected": -1.0144208669662476, |
| "step": 81, |
| "train_speed(iter/s)": 0.010699 |
| }, |
| { |
| "epoch": 0.18298465829846583, |
| "grad_norm": 1.6746541261672974, |
| "learning_rate": 0.00019906859460363307, |
| "logits/chosen": -0.4001500606536865, |
| "logits/rejected": -0.45929989218711853, |
| "logps/chosen": -7.395153045654297, |
| "logps/rejected": -25.426557540893555, |
| "loss": 0.7357066869735718, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4178068935871124, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.16341763734817505, |
| "rewards/margins": 1.525546669960022, |
| "rewards/rejected": -1.3621290922164917, |
| "step": 82, |
| "train_speed(iter/s)": 0.0107 |
| }, |
| { |
| "epoch": 0.18521617852161784, |
| "grad_norm": 4.773955821990967, |
| "learning_rate": 0.00019901765171489086, |
| "logits/chosen": -0.30332040786743164, |
| "logits/rejected": -0.4577845633029938, |
| "logps/chosen": -10.913299560546875, |
| "logps/rejected": -46.28175354003906, |
| "loss": 1.1305065155029297, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.7709158658981323, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.34224939346313477, |
| "rewards/margins": 2.636122703552246, |
| "rewards/rejected": -2.9783718585968018, |
| "step": 83, |
| "train_speed(iter/s)": 0.0107 |
| }, |
| { |
| "epoch": 0.18744769874476988, |
| "grad_norm": 2.5695321559906006, |
| "learning_rate": 0.0001989653593887117, |
| "logits/chosen": -0.3644600212574005, |
| "logits/rejected": -0.43475809693336487, |
| "logps/chosen": -6.348275661468506, |
| "logps/rejected": -29.636547088623047, |
| "loss": 0.8502273559570312, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4918310344219208, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07552118599414825, |
| "rewards/margins": 1.7124391794204712, |
| "rewards/rejected": -1.636918067932129, |
| "step": 84, |
| "train_speed(iter/s)": 0.010702 |
| }, |
| { |
| "epoch": 0.1896792189679219, |
| "grad_norm": 3.117232322692871, |
| "learning_rate": 0.00019891171833774854, |
| "logits/chosen": -0.31497788429260254, |
| "logits/rejected": -0.40428590774536133, |
| "logps/chosen": -10.195795059204102, |
| "logps/rejected": -43.74475860595703, |
| "loss": 1.0120048522949219, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6030641198158264, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.17749960720539093, |
| "rewards/margins": 2.177133083343506, |
| "rewards/rejected": -2.35463285446167, |
| "step": 85, |
| "train_speed(iter/s)": 0.010701 |
| }, |
| { |
| "epoch": 0.19191073919107393, |
| "grad_norm": 2.5167582035064697, |
| "learning_rate": 0.00019885672929303508, |
| "logits/chosen": -0.3179773986339569, |
| "logits/rejected": -0.42796629667282104, |
| "logps/chosen": -10.96944522857666, |
| "logps/rejected": -40.80889129638672, |
| "loss": 0.8473846912384033, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5096143484115601, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.08502935618162155, |
| "rewards/margins": 2.197641134262085, |
| "rewards/rejected": -2.282670259475708, |
| "step": 86, |
| "train_speed(iter/s)": 0.010702 |
| }, |
| { |
| "epoch": 0.19414225941422594, |
| "grad_norm": 2.2492237091064453, |
| "learning_rate": 0.00019880039300397591, |
| "logits/chosen": -0.4064781665802002, |
| "logits/rejected": -0.4387715458869934, |
| "logps/chosen": -7.350955009460449, |
| "logps/rejected": -30.075700759887695, |
| "loss": 0.9061151742935181, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5411654710769653, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.04652661085128784, |
| "rewards/margins": 1.5254417657852173, |
| "rewards/rejected": -1.4789149761199951, |
| "step": 87, |
| "train_speed(iter/s)": 0.010703 |
| }, |
| { |
| "epoch": 0.19637377963737795, |
| "grad_norm": 1.698593258857727, |
| "learning_rate": 0.00019874271023833604, |
| "logits/chosen": -0.36257320642471313, |
| "logits/rejected": -0.4434874653816223, |
| "logps/chosen": -8.048574447631836, |
| "logps/rejected": -30.737728118896484, |
| "loss": 0.8314277529716492, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.47138625383377075, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.10594448447227478, |
| "rewards/margins": 1.524275541305542, |
| "rewards/rejected": -1.4183311462402344, |
| "step": 88, |
| "train_speed(iter/s)": 0.010703 |
| }, |
| { |
| "epoch": 0.19860529986053, |
| "grad_norm": 3.0257344245910645, |
| "learning_rate": 0.00019868368178223075, |
| "logits/chosen": -0.3720957636833191, |
| "logits/rejected": -0.45681309700012207, |
| "logps/chosen": -9.780877113342285, |
| "logps/rejected": -23.92654037475586, |
| "loss": 1.1227757930755615, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6208180785179138, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.049955643713474274, |
| "rewards/margins": 0.7902018427848816, |
| "rewards/rejected": -0.8401575088500977, |
| "step": 89, |
| "train_speed(iter/s)": 0.010704 |
| }, |
| { |
| "epoch": 0.200836820083682, |
| "grad_norm": 1.996423363685608, |
| "learning_rate": 0.00019862330844011466, |
| "logits/chosen": -0.4094502031803131, |
| "logits/rejected": -0.5237057209014893, |
| "logps/chosen": -7.119333744049072, |
| "logps/rejected": -27.10557746887207, |
| "loss": 0.9298258423805237, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5608960390090942, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13814933598041534, |
| "rewards/margins": 1.2807323932647705, |
| "rewards/rejected": -1.1425830125808716, |
| "step": 90, |
| "train_speed(iter/s)": 0.010705 |
| }, |
| { |
| "epoch": 0.20306834030683404, |
| "grad_norm": 2.4107954502105713, |
| "learning_rate": 0.00019856159103477086, |
| "logits/chosen": -0.38746166229248047, |
| "logits/rejected": -0.488497257232666, |
| "logps/chosen": -4.5973896980285645, |
| "logps/rejected": -27.419519424438477, |
| "loss": 0.710839033126831, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.30818992853164673, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.038327932357788086, |
| "rewards/margins": 1.1028857231140137, |
| "rewards/rejected": -1.064557671546936, |
| "step": 91, |
| "train_speed(iter/s)": 0.010705 |
| }, |
| { |
| "epoch": 0.20529986052998606, |
| "grad_norm": 1.7134486436843872, |
| "learning_rate": 0.00019849853040729962, |
| "logits/chosen": -0.39711794257164, |
| "logits/rejected": -0.5056676268577576, |
| "logps/chosen": -3.698580265045166, |
| "logps/rejected": -35.89771270751953, |
| "loss": 0.6057847142219543, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.28385108709335327, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.012411292642354965, |
| "rewards/margins": 1.7848464250564575, |
| "rewards/rejected": -1.7724350690841675, |
| "step": 92, |
| "train_speed(iter/s)": 0.010705 |
| }, |
| { |
| "epoch": 0.20753138075313807, |
| "grad_norm": 3.236150026321411, |
| "learning_rate": 0.00019843412741710705, |
| "logits/chosen": -0.4137122631072998, |
| "logits/rejected": -0.5312527418136597, |
| "logps/chosen": -8.622723579406738, |
| "logps/rejected": -32.65446853637695, |
| "loss": 0.923988938331604, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5595443248748779, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.008658979088068008, |
| "rewards/margins": 1.6513465642929077, |
| "rewards/rejected": -1.6600055694580078, |
| "step": 93, |
| "train_speed(iter/s)": 0.010706 |
| }, |
| { |
| "epoch": 0.2097629009762901, |
| "grad_norm": 3.0684919357299805, |
| "learning_rate": 0.00019836838294189327, |
| "logits/chosen": -0.30109331011772156, |
| "logits/rejected": -0.5258408784866333, |
| "logps/chosen": -4.7982635498046875, |
| "logps/rejected": -46.58249282836914, |
| "loss": 0.6704127788543701, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.41316378116607666, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.022386539727449417, |
| "rewards/margins": 2.3209354877471924, |
| "rewards/rejected": -2.3433220386505127, |
| "step": 94, |
| "train_speed(iter/s)": 0.010705 |
| }, |
| { |
| "epoch": 0.21199442119944212, |
| "grad_norm": 1.8835054636001587, |
| "learning_rate": 0.00019830129787764048, |
| "logits/chosen": -0.41429489850997925, |
| "logits/rejected": -0.5583072304725647, |
| "logps/chosen": -7.361013412475586, |
| "logps/rejected": -39.35795211791992, |
| "loss": 0.6725447773933411, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.368007630109787, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.1507667899131775, |
| "rewards/margins": 2.0595011711120605, |
| "rewards/rejected": -2.210268020629883, |
| "step": 95, |
| "train_speed(iter/s)": 0.010706 |
| }, |
| { |
| "epoch": 0.21422594142259413, |
| "grad_norm": 2.3120245933532715, |
| "learning_rate": 0.00019823287313860087, |
| "logits/chosen": -0.372950941324234, |
| "logits/rejected": -0.49759799242019653, |
| "logps/chosen": -11.2495756149292, |
| "logps/rejected": -34.734066009521484, |
| "loss": 1.1070480346679688, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6910135746002197, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.07585638761520386, |
| "rewards/margins": 1.6244981288909912, |
| "rewards/rejected": -1.7003545761108398, |
| "step": 96, |
| "train_speed(iter/s)": 0.010706 |
| }, |
| { |
| "epoch": 0.21645746164574617, |
| "grad_norm": 2.277174711227417, |
| "learning_rate": 0.00019816310965728388, |
| "logits/chosen": -0.4299129247665405, |
| "logits/rejected": -0.4980982542037964, |
| "logps/chosen": -9.542954444885254, |
| "logps/rejected": -25.756847381591797, |
| "loss": 0.9643784165382385, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4663383364677429, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.04499448090791702, |
| "rewards/margins": 1.144527554512024, |
| "rewards/rejected": -1.1895220279693604, |
| "step": 97, |
| "train_speed(iter/s)": 0.010706 |
| }, |
| { |
| "epoch": 0.21868898186889818, |
| "grad_norm": 1.9245892763137817, |
| "learning_rate": 0.00019809200838444383, |
| "logits/chosen": -0.4750334322452545, |
| "logits/rejected": -0.5499654412269592, |
| "logps/chosen": -7.279554843902588, |
| "logps/rejected": -27.49015998840332, |
| "loss": 0.7538543343544006, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.37693145871162415, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.19298376142978668, |
| "rewards/margins": 1.2535287141799927, |
| "rewards/rejected": -1.0605449676513672, |
| "step": 98, |
| "train_speed(iter/s)": 0.010707 |
| }, |
| { |
| "epoch": 0.22092050209205022, |
| "grad_norm": 2.6517419815063477, |
| "learning_rate": 0.0001980195702890667, |
| "logits/chosen": -0.3958469033241272, |
| "logits/rejected": -0.5089414715766907, |
| "logps/chosen": -7.152680397033691, |
| "logps/rejected": -34.461917877197266, |
| "loss": 0.8412438035011292, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3738154470920563, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.022486146539449692, |
| "rewards/margins": 1.67362380027771, |
| "rewards/rejected": -1.6961098909378052, |
| "step": 99, |
| "train_speed(iter/s)": 0.010708 |
| }, |
| { |
| "epoch": 0.22315202231520223, |
| "grad_norm": 2.8246142864227295, |
| "learning_rate": 0.00019794579635835704, |
| "logits/chosen": -0.4292357861995697, |
| "logits/rejected": -0.5326910614967346, |
| "logps/chosen": -7.5378594398498535, |
| "logps/rejected": -21.57219696044922, |
| "loss": 1.0386667251586914, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5401602387428284, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.06580009311437607, |
| "rewards/margins": 0.625900149345398, |
| "rewards/rejected": -0.5601000189781189, |
| "step": 100, |
| "train_speed(iter/s)": 0.010708 |
| }, |
| { |
| "epoch": 0.22538354253835424, |
| "grad_norm": 2.5541062355041504, |
| "learning_rate": 0.00019787068759772458, |
| "logits/chosen": -0.2897275686264038, |
| "logits/rejected": -0.469921737909317, |
| "logps/chosen": -5.731183052062988, |
| "logps/rejected": -42.3974494934082, |
| "loss": 0.6600039005279541, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.34703782200813293, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.03177756816148758, |
| "rewards/margins": 2.005840539932251, |
| "rewards/rejected": -1.9740626811981201, |
| "step": 101, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.22761506276150628, |
| "grad_norm": 1.7807402610778809, |
| "learning_rate": 0.0001977942450307703, |
| "logits/chosen": -0.4453420042991638, |
| "logits/rejected": -0.5632447004318237, |
| "logps/chosen": -2.4129638671875, |
| "logps/rejected": -25.95719337463379, |
| "loss": 0.5793961882591248, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.23024141788482666, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.14239919185638428, |
| "rewards/margins": 1.4281731843948364, |
| "rewards/rejected": -1.2857739925384521, |
| "step": 102, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.2298465829846583, |
| "grad_norm": 2.2132644653320312, |
| "learning_rate": 0.00019771646969927276, |
| "logits/chosen": -0.35926875472068787, |
| "logits/rejected": -0.520535945892334, |
| "logps/chosen": -13.426995277404785, |
| "logps/rejected": -45.62168884277344, |
| "loss": 1.015347957611084, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.7125861644744873, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.29009538888931274, |
| "rewards/margins": 2.0907297134399414, |
| "rewards/rejected": -2.3808252811431885, |
| "step": 103, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.23207810320781033, |
| "grad_norm": 3.2295114994049072, |
| "learning_rate": 0.00019763736266317373, |
| "logits/chosen": -0.40255796909332275, |
| "logits/rejected": -0.48326143622398376, |
| "logps/chosen": -11.606037139892578, |
| "logps/rejected": -31.71380043029785, |
| "loss": 1.1166499853134155, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6979129910469055, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.08108239620923996, |
| "rewards/margins": 1.4596813917160034, |
| "rewards/rejected": -1.5407637357711792, |
| "step": 104, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.23430962343096234, |
| "grad_norm": 10.526008605957031, |
| "learning_rate": 0.00019755692500056377, |
| "logits/chosen": -0.39488697052001953, |
| "logits/rejected": -0.4993595480918884, |
| "logps/chosen": -8.515541076660156, |
| "logps/rejected": -43.509639739990234, |
| "loss": 0.9588133096694946, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6071394085884094, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.21072883903980255, |
| "rewards/margins": 2.291653633117676, |
| "rewards/rejected": -2.502382278442383, |
| "step": 105, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.23654114365411436, |
| "grad_norm": 4.369258880615234, |
| "learning_rate": 0.0001974751578076675, |
| "logits/chosen": -0.47596731781959534, |
| "logits/rejected": -0.567024827003479, |
| "logps/chosen": -9.769340515136719, |
| "logps/rejected": -36.51396560668945, |
| "loss": 1.1472177505493164, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.7814697623252869, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.07862536609172821, |
| "rewards/margins": 1.9599454402923584, |
| "rewards/rejected": -2.0385708808898926, |
| "step": 106, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.2387726638772664, |
| "grad_norm": 2.6377806663513184, |
| "learning_rate": 0.0001973920621988288, |
| "logits/chosen": -0.4363410472869873, |
| "logits/rejected": -0.5180681943893433, |
| "logps/chosen": -8.525186538696289, |
| "logps/rejected": -36.119422912597656, |
| "loss": 0.8840208053588867, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5660971999168396, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.11222146451473236, |
| "rewards/margins": 2.031522035598755, |
| "rewards/rejected": -2.1437435150146484, |
| "step": 107, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.2410041841004184, |
| "grad_norm": 3.5556135177612305, |
| "learning_rate": 0.00019730763930649557, |
| "logits/chosen": -0.4549260139465332, |
| "logits/rejected": -0.5709996223449707, |
| "logps/chosen": -6.807540416717529, |
| "logps/rejected": -35.386817932128906, |
| "loss": 1.0365853309631348, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6718467473983765, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.010064257308840752, |
| "rewards/margins": 2.010096549987793, |
| "rewards/rejected": -2.0201611518859863, |
| "step": 108, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.24323570432357045, |
| "grad_norm": 2.9586894512176514, |
| "learning_rate": 0.0001972218902812041, |
| "logits/chosen": -0.5169477462768555, |
| "logits/rejected": -0.5766913890838623, |
| "logps/chosen": -9.75178050994873, |
| "logps/rejected": -26.94969940185547, |
| "loss": 1.03122079372406, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6585339307785034, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.047493062913417816, |
| "rewards/margins": 1.4683667421340942, |
| "rewards/rejected": -1.5158597230911255, |
| "step": 109, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.24546722454672246, |
| "grad_norm": 2.40643310546875, |
| "learning_rate": 0.0001971348162915637, |
| "logits/chosen": -0.40012210607528687, |
| "logits/rejected": -0.5228334665298462, |
| "logps/chosen": -12.064148902893066, |
| "logps/rejected": -42.65177917480469, |
| "loss": 1.028669834136963, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.584478497505188, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.0045367032289505005, |
| "rewards/margins": 1.1938111782073975, |
| "rewards/rejected": -1.189274549484253, |
| "step": 110, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.24769874476987447, |
| "grad_norm": 1.9528326988220215, |
| "learning_rate": 0.0001970464185242406, |
| "logits/chosen": -0.48454731702804565, |
| "logits/rejected": -0.5713267922401428, |
| "logps/chosen": -8.57382869720459, |
| "logps/rejected": -28.3697452545166, |
| "loss": 0.9442793130874634, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5884312987327576, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": -0.007765632122755051, |
| "rewards/margins": 1.1733622550964355, |
| "rewards/rejected": -1.181127905845642, |
| "step": 111, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.2499302649930265, |
| "grad_norm": 1.6935343742370605, |
| "learning_rate": 0.00019695669818394176, |
| "logits/chosen": -0.49247032403945923, |
| "logits/rejected": -0.5529605150222778, |
| "logps/chosen": -6.149413108825684, |
| "logps/rejected": -26.09927749633789, |
| "loss": 0.8117169141769409, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.36330851912498474, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.11475833505392075, |
| "rewards/margins": 0.9903795719146729, |
| "rewards/rejected": -0.8756212592124939, |
| "step": 112, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.25216178521617855, |
| "grad_norm": 1.7400895357131958, |
| "learning_rate": 0.00019686565649339852, |
| "logits/chosen": -0.47781193256378174, |
| "logits/rejected": -0.541789174079895, |
| "logps/chosen": -7.246283054351807, |
| "logps/rejected": -28.977123260498047, |
| "loss": 0.8169933557510376, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.46227994561195374, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1930399090051651, |
| "rewards/margins": 1.5713894367218018, |
| "rewards/rejected": -1.3783494234085083, |
| "step": 113, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.25439330543933053, |
| "grad_norm": 2.0717272758483887, |
| "learning_rate": 0.0001967732946933499, |
| "logits/chosen": -0.48887836933135986, |
| "logits/rejected": -0.5135952234268188, |
| "logps/chosen": -9.175788879394531, |
| "logps/rejected": -19.074743270874023, |
| "loss": 1.0894757509231567, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5358699560165405, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.011165066622197628, |
| "rewards/margins": 0.7140389084815979, |
| "rewards/rejected": -0.7252039313316345, |
| "step": 114, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.25662482566248257, |
| "grad_norm": 1.8609806299209595, |
| "learning_rate": 0.00019667961404252573, |
| "logits/chosen": -0.46024733781814575, |
| "logits/rejected": -0.5330649018287659, |
| "logps/chosen": -6.905147552490234, |
| "logps/rejected": -30.54239273071289, |
| "loss": 0.8680601119995117, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.44586634635925293, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.03578485921025276, |
| "rewards/margins": 1.5293731689453125, |
| "rewards/rejected": -1.4935882091522217, |
| "step": 115, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.2588563458856346, |
| "grad_norm": 1.66657292842865, |
| "learning_rate": 0.00019658461581762948, |
| "logits/chosen": -0.430117666721344, |
| "logits/rejected": -0.5585033297538757, |
| "logps/chosen": -5.328842639923096, |
| "logps/rejected": -37.261329650878906, |
| "loss": 0.6072223782539368, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3235556483268738, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.11161011457443237, |
| "rewards/margins": 2.0768401622772217, |
| "rewards/rejected": -1.965229868888855, |
| "step": 116, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.2610878661087866, |
| "grad_norm": 1.51878821849823, |
| "learning_rate": 0.0001964883013133208, |
| "logits/chosen": -0.419185996055603, |
| "logits/rejected": -0.5297868251800537, |
| "logps/chosen": -8.04638385772705, |
| "logps/rejected": -37.03801727294922, |
| "loss": 0.7135862708091736, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.40418556332588196, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.003706634044647217, |
| "rewards/margins": 2.0922625064849854, |
| "rewards/rejected": -2.0959692001342773, |
| "step": 117, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.26331938633193863, |
| "grad_norm": 1.7501200437545776, |
| "learning_rate": 0.00019639067184219796, |
| "logits/chosen": -0.4789353609085083, |
| "logits/rejected": -0.5624681711196899, |
| "logps/chosen": -2.698453426361084, |
| "logps/rejected": -30.66053009033203, |
| "loss": 0.4839392304420471, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.23937827348709106, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1991066038608551, |
| "rewards/margins": 2.1478512287139893, |
| "rewards/rejected": -1.948744773864746, |
| "step": 118, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.26555090655509067, |
| "grad_norm": 5.462699890136719, |
| "learning_rate": 0.00019629172873477995, |
| "logits/chosen": -0.4896751642227173, |
| "logits/rejected": -0.6056660413742065, |
| "logps/chosen": -9.186168670654297, |
| "logps/rejected": -43.67456817626953, |
| "loss": 1.0974973440170288, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.7505476474761963, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.26940032839775085, |
| "rewards/margins": 2.277892589569092, |
| "rewards/rejected": -2.547292709350586, |
| "step": 119, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.26778242677824265, |
| "grad_norm": 2.521127462387085, |
| "learning_rate": 0.00019619147333948823, |
| "logits/chosen": -0.40272024273872375, |
| "logits/rejected": -0.6465494632720947, |
| "logps/chosen": -4.102967262268066, |
| "logps/rejected": -46.14149475097656, |
| "loss": 0.6112779974937439, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.35009804368019104, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.06559442728757858, |
| "rewards/margins": 2.652864456176758, |
| "rewards/rejected": -2.718458652496338, |
| "step": 120, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.2700139470013947, |
| "grad_norm": 6.966860771179199, |
| "learning_rate": 0.00019608990702262862, |
| "logits/chosen": -0.46823495626449585, |
| "logits/rejected": -0.5669148564338684, |
| "logps/chosen": -11.427492141723633, |
| "logps/rejected": -41.708621978759766, |
| "loss": 0.9116367101669312, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5967401266098022, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.1262148916721344, |
| "rewards/margins": 2.7404427528381348, |
| "rewards/rejected": -2.8666577339172363, |
| "step": 121, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.27224546722454673, |
| "grad_norm": 3.0951266288757324, |
| "learning_rate": 0.0001959870311683723, |
| "logits/chosen": -0.4486353099346161, |
| "logits/rejected": -0.5479130744934082, |
| "logps/chosen": -12.51127815246582, |
| "logps/rejected": -38.43931198120117, |
| "loss": 0.9770362377166748, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5739753246307373, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.2812083661556244, |
| "rewards/margins": 1.7542171478271484, |
| "rewards/rejected": -2.0354256629943848, |
| "step": 122, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.2744769874476987, |
| "grad_norm": 2.664623498916626, |
| "learning_rate": 0.00019588284717873738, |
| "logits/chosen": -0.48905450105667114, |
| "logits/rejected": -0.5784589052200317, |
| "logps/chosen": -11.445174217224121, |
| "logps/rejected": -45.63172149658203, |
| "loss": 0.8187350630760193, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5693225860595703, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.07134091854095459, |
| "rewards/margins": 2.74631929397583, |
| "rewards/rejected": -2.8176605701446533, |
| "step": 123, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.27670850767085076, |
| "grad_norm": 1.5270544290542603, |
| "learning_rate": 0.00019577735647356928, |
| "logits/chosen": -0.42984533309936523, |
| "logits/rejected": -0.5359001159667969, |
| "logps/chosen": -6.653199195861816, |
| "logps/rejected": -29.57577133178711, |
| "loss": 0.7354912757873535, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3922813832759857, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1660708487033844, |
| "rewards/margins": 1.5390381813049316, |
| "rewards/rejected": -1.3729673624038696, |
| "step": 124, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.2789400278940028, |
| "grad_norm": 3.5429413318634033, |
| "learning_rate": 0.00019567056049052203, |
| "logits/chosen": -0.3107150197029114, |
| "logits/rejected": -0.4991438388824463, |
| "logps/chosen": -5.079807758331299, |
| "logps/rejected": -40.358970642089844, |
| "loss": 0.6619258522987366, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.36486440896987915, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.030078772455453873, |
| "rewards/margins": 1.9852522611618042, |
| "rewards/rejected": -1.955173373222351, |
| "step": 125, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.28117154811715483, |
| "grad_norm": 1.961870551109314, |
| "learning_rate": 0.00019556246068503795, |
| "logits/chosen": -0.44410189986228943, |
| "logits/rejected": -0.4920841455459595, |
| "logps/chosen": -11.91314697265625, |
| "logps/rejected": -31.1485595703125, |
| "loss": 0.8891083598136902, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5505969524383545, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.03779969736933708, |
| "rewards/margins": 1.8280887603759766, |
| "rewards/rejected": -1.790289044380188, |
| "step": 126, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.2834030683403068, |
| "grad_norm": 2.261699676513672, |
| "learning_rate": 0.00019545305853032848, |
| "logits/chosen": -0.3864215016365051, |
| "logits/rejected": -0.4863501191139221, |
| "logps/chosen": -7.286771774291992, |
| "logps/rejected": -30.703035354614258, |
| "loss": 0.8944107294082642, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5101214051246643, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.018673699349164963, |
| "rewards/margins": 1.4196372032165527, |
| "rewards/rejected": -1.400963544845581, |
| "step": 127, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.28563458856345886, |
| "grad_norm": 1.475645899772644, |
| "learning_rate": 0.0001953423555173536, |
| "logits/chosen": -0.36042094230651855, |
| "logits/rejected": -0.5007886290550232, |
| "logps/chosen": -5.484119892120361, |
| "logps/rejected": -41.732566833496094, |
| "loss": 0.587951123714447, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.33234262466430664, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.022678524255752563, |
| "rewards/margins": 1.7825675010681152, |
| "rewards/rejected": -1.7598888874053955, |
| "step": 128, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.2878661087866109, |
| "grad_norm": 1.8218269348144531, |
| "learning_rate": 0.0001952303531548018, |
| "logits/chosen": -0.33974117040634155, |
| "logits/rejected": -0.44949695467948914, |
| "logps/chosen": -5.385599136352539, |
| "logps/rejected": -34.09246063232422, |
| "loss": 0.566850483417511, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.25613701343536377, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.12256469577550888, |
| "rewards/margins": 1.9903541803359985, |
| "rewards/rejected": -1.8677895069122314, |
| "step": 129, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.2900976290097629, |
| "grad_norm": 1.8668869733810425, |
| "learning_rate": 0.00019511705296906945, |
| "logits/chosen": -0.3487342596054077, |
| "logits/rejected": -0.4806521534919739, |
| "logps/chosen": -4.425655841827393, |
| "logps/rejected": -30.515653610229492, |
| "loss": 0.5262651443481445, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.25739917159080505, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.12762486934661865, |
| "rewards/margins": 1.6179413795471191, |
| "rewards/rejected": -1.4903165102005005, |
| "step": 130, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.2923291492329149, |
| "grad_norm": 2.4468886852264404, |
| "learning_rate": 0.00019500245650423987, |
| "logits/chosen": -0.3514082133769989, |
| "logits/rejected": -0.4678734242916107, |
| "logps/chosen": -9.789356231689453, |
| "logps/rejected": -43.26856231689453, |
| "loss": 0.8514004349708557, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5882378816604614, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.00511053204536438, |
| "rewards/margins": 2.595022678375244, |
| "rewards/rejected": -2.600132942199707, |
| "step": 131, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.29456066945606696, |
| "grad_norm": 1.4899861812591553, |
| "learning_rate": 0.00019488656532206256, |
| "logits/chosen": -0.32015788555145264, |
| "logits/rejected": -0.5057851672172546, |
| "logps/chosen": -5.5769572257995605, |
| "logps/rejected": -39.262237548828125, |
| "loss": 0.5258051156997681, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.2577199935913086, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.012265877798199654, |
| "rewards/margins": 1.970078945159912, |
| "rewards/rejected": -1.957813024520874, |
| "step": 132, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.29679218967921894, |
| "grad_norm": 2.7044637203216553, |
| "learning_rate": 0.00019476938100193166, |
| "logits/chosen": -0.3656051754951477, |
| "logits/rejected": -0.4236447215080261, |
| "logps/chosen": -10.65760612487793, |
| "logps/rejected": -38.63339614868164, |
| "loss": 0.9246849417686462, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6106225848197937, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.030058247968554497, |
| "rewards/margins": 2.29494047164917, |
| "rewards/rejected": -2.2648820877075195, |
| "step": 133, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.299023709902371, |
| "grad_norm": 1.82398521900177, |
| "learning_rate": 0.00019465090514086448, |
| "logits/chosen": -0.2924862504005432, |
| "logits/rejected": -0.49178171157836914, |
| "logps/chosen": -5.639602184295654, |
| "logps/rejected": -44.825313568115234, |
| "loss": 0.5453966856002808, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3345707952976227, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07429855316877365, |
| "rewards/margins": 2.4938178062438965, |
| "rewards/rejected": -2.4195189476013184, |
| "step": 134, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.301255230125523, |
| "grad_norm": 3.8286569118499756, |
| "learning_rate": 0.00019453113935347983, |
| "logits/chosen": -0.29531559348106384, |
| "logits/rejected": -0.43998682498931885, |
| "logps/chosen": -10.71696662902832, |
| "logps/rejected": -31.618389129638672, |
| "loss": 1.1089136600494385, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.7216206789016724, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.1611023098230362, |
| "rewards/margins": 1.5489929914474487, |
| "rewards/rejected": -1.7100951671600342, |
| "step": 135, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.30348675034867506, |
| "grad_norm": 2.3694629669189453, |
| "learning_rate": 0.000194410085271976, |
| "logits/chosen": -0.3209247291088104, |
| "logits/rejected": -0.47681719064712524, |
| "logps/chosen": -8.806313514709473, |
| "logps/rejected": -42.08256912231445, |
| "loss": 0.8723977208137512, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6001561880111694, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.08420376479625702, |
| "rewards/margins": 2.307250499725342, |
| "rewards/rejected": -2.3914542198181152, |
| "step": 136, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.30571827057182704, |
| "grad_norm": 4.322510719299316, |
| "learning_rate": 0.00019428774454610843, |
| "logits/chosen": -0.3234959542751312, |
| "logits/rejected": -0.43433526158332825, |
| "logps/chosen": -8.827983856201172, |
| "logps/rejected": -38.56624984741211, |
| "loss": 0.8363085985183716, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5304505825042725, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.03617179021239281, |
| "rewards/margins": 2.3936820030212402, |
| "rewards/rejected": -2.429853677749634, |
| "step": 137, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.3079497907949791, |
| "grad_norm": 1.8575319051742554, |
| "learning_rate": 0.00019416411884316722, |
| "logits/chosen": -0.19573748111724854, |
| "logits/rejected": -0.39096614718437195, |
| "logps/chosen": -5.302525043487549, |
| "logps/rejected": -39.366905212402344, |
| "loss": 0.5817482471466064, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3171233534812927, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.06202463060617447, |
| "rewards/margins": 2.1402504444122314, |
| "rewards/rejected": -2.07822585105896, |
| "step": 138, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.3101813110181311, |
| "grad_norm": 1.6324138641357422, |
| "learning_rate": 0.00019403920984795454, |
| "logits/chosen": -0.18411299586296082, |
| "logits/rejected": -0.3911224901676178, |
| "logps/chosen": -4.253074645996094, |
| "logps/rejected": -37.889976501464844, |
| "loss": 0.6420720815658569, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3620051145553589, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1460508555173874, |
| "rewards/margins": 2.2889294624328613, |
| "rewards/rejected": -2.142878770828247, |
| "step": 139, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.3124128312412831, |
| "grad_norm": 2.665651798248291, |
| "learning_rate": 0.00019391301926276156, |
| "logits/chosen": -0.22396986186504364, |
| "logits/rejected": -0.3946935534477234, |
| "logps/chosen": -4.665857315063477, |
| "logps/rejected": -40.48711395263672, |
| "loss": 0.7322566509246826, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.41601666808128357, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.039280496537685394, |
| "rewards/margins": 2.420466899871826, |
| "rewards/rejected": -2.459747314453125, |
| "step": 140, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.31464435146443515, |
| "grad_norm": 2.0447568893432617, |
| "learning_rate": 0.00019378554880734527, |
| "logits/chosen": -0.22006088495254517, |
| "logits/rejected": -0.3875121772289276, |
| "logps/chosen": -5.640967845916748, |
| "logps/rejected": -47.2126579284668, |
| "loss": 0.6299532651901245, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4075142443180084, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.07418311387300491, |
| "rewards/margins": 3.378464460372925, |
| "rewards/rejected": -3.304281711578369, |
| "step": 141, |
| "train_speed(iter/s)": 0.010695 |
| }, |
| { |
| "epoch": 0.3168758716875872, |
| "grad_norm": 3.150909185409546, |
| "learning_rate": 0.00019365680021890506, |
| "logits/chosen": -0.2578600347042084, |
| "logits/rejected": -0.35254308581352234, |
| "logps/chosen": -8.539968490600586, |
| "logps/rejected": -34.56965637207031, |
| "loss": 0.9595645070075989, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5309237241744995, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.1485617607831955, |
| "rewards/margins": 1.7760200500488281, |
| "rewards/rejected": -1.9245820045471191, |
| "step": 142, |
| "train_speed(iter/s)": 0.010696 |
| }, |
| { |
| "epoch": 0.31910739191073917, |
| "grad_norm": 3.2599754333496094, |
| "learning_rate": 0.0001935267752520591, |
| "logits/chosen": -0.3178224563598633, |
| "logits/rejected": -0.43051257729530334, |
| "logps/chosen": -9.458460807800293, |
| "logps/rejected": -43.98479080200195, |
| "loss": 0.8238604068756104, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5105634331703186, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.19997264444828033, |
| "rewards/margins": 2.638275146484375, |
| "rewards/rejected": -2.8382480144500732, |
| "step": 143, |
| "train_speed(iter/s)": 0.010696 |
| }, |
| { |
| "epoch": 0.3213389121338912, |
| "grad_norm": 7.584465503692627, |
| "learning_rate": 0.00019339547567882024, |
| "logits/chosen": -0.2016015499830246, |
| "logits/rejected": -0.36200886964797974, |
| "logps/chosen": -9.776619911193848, |
| "logps/rejected": -48.03835678100586, |
| "loss": 0.9604641199111938, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.7163732647895813, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.24650944769382477, |
| "rewards/margins": 3.3983724117279053, |
| "rewards/rejected": -3.151862621307373, |
| "step": 144, |
| "train_speed(iter/s)": 0.010696 |
| }, |
| { |
| "epoch": 0.32357043235704325, |
| "grad_norm": 3.555295944213867, |
| "learning_rate": 0.0001932629032885721, |
| "logits/chosen": -0.26102834939956665, |
| "logits/rejected": -0.3585357069969177, |
| "logps/chosen": -12.133922576904297, |
| "logps/rejected": -48.155155181884766, |
| "loss": 0.9948782920837402, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6610950231552124, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.4001372456550598, |
| "rewards/margins": 2.751622200012207, |
| "rewards/rejected": -3.151759386062622, |
| "step": 145, |
| "train_speed(iter/s)": 0.010697 |
| }, |
| { |
| "epoch": 0.32580195258019523, |
| "grad_norm": 2.2978155612945557, |
| "learning_rate": 0.0001931290598880445, |
| "logits/chosen": -0.20670001208782196, |
| "logits/rejected": -0.38169875741004944, |
| "logps/chosen": -6.4796600341796875, |
| "logps/rejected": -41.426429748535156, |
| "loss": 0.6576346158981323, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.42290252447128296, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.11042710393667221, |
| "rewards/margins": 2.6782710552215576, |
| "rewards/rejected": -2.5678436756134033, |
| "step": 146, |
| "train_speed(iter/s)": 0.010697 |
| }, |
| { |
| "epoch": 0.32803347280334727, |
| "grad_norm": 2.1882224082946777, |
| "learning_rate": 0.00019299394730128895, |
| "logits/chosen": -0.20903249084949493, |
| "logits/rejected": -0.2923564314842224, |
| "logps/chosen": -12.293680191040039, |
| "logps/rejected": -50.68069076538086, |
| "loss": 0.8728819489479065, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5617231130599976, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.08868011832237244, |
| "rewards/margins": 3.250684976577759, |
| "rewards/rejected": -3.339365005493164, |
| "step": 147, |
| "train_speed(iter/s)": 0.010696 |
| }, |
| { |
| "epoch": 0.3302649930264993, |
| "grad_norm": 2.5009729862213135, |
| "learning_rate": 0.00019285756736965367, |
| "logits/chosen": -0.2492772340774536, |
| "logits/rejected": -0.3531278669834137, |
| "logps/chosen": -7.951318264007568, |
| "logps/rejected": -34.95103073120117, |
| "loss": 0.9188458323478699, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5677845478057861, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.015583137050271034, |
| "rewards/margins": 2.127014398574829, |
| "rewards/rejected": -2.111431360244751, |
| "step": 148, |
| "train_speed(iter/s)": 0.010697 |
| }, |
| { |
| "epoch": 0.33249651324965135, |
| "grad_norm": 2.460752487182617, |
| "learning_rate": 0.00019271992195175875, |
| "logits/chosen": -0.31061843037605286, |
| "logits/rejected": -0.42540180683135986, |
| "logps/chosen": -8.994503021240234, |
| "logps/rejected": -38.728179931640625, |
| "loss": 0.9080746173858643, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5827285051345825, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.09331205487251282, |
| "rewards/margins": 2.394498348236084, |
| "rewards/rejected": -2.3011860847473145, |
| "step": 149, |
| "train_speed(iter/s)": 0.010697 |
| }, |
| { |
| "epoch": 0.33472803347280333, |
| "grad_norm": 1.7428467273712158, |
| "learning_rate": 0.00019258101292347042, |
| "logits/chosen": -0.19874219596385956, |
| "logits/rejected": -0.36566510796546936, |
| "logps/chosen": -8.282856941223145, |
| "logps/rejected": -47.80164337158203, |
| "loss": 0.6850791573524475, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5025182366371155, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.09450387954711914, |
| "rewards/margins": 2.836395502090454, |
| "rewards/rejected": -2.741891622543335, |
| "step": 150, |
| "train_speed(iter/s)": 0.010697 |
| }, |
| { |
| "epoch": 0.33695955369595537, |
| "grad_norm": 1.7387069463729858, |
| "learning_rate": 0.00019244084217787588, |
| "logits/chosen": -0.22253645956516266, |
| "logits/rejected": -0.36506688594818115, |
| "logps/chosen": -8.663468360900879, |
| "logps/rejected": -48.536319732666016, |
| "loss": 0.7012121081352234, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.469788521528244, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.1591811329126358, |
| "rewards/margins": 2.9181699752807617, |
| "rewards/rejected": -2.758988857269287, |
| "step": 151, |
| "train_speed(iter/s)": 0.010684 |
| }, |
| { |
| "epoch": 0.3391910739191074, |
| "grad_norm": 2.085836887359619, |
| "learning_rate": 0.00019229941162525726, |
| "logits/chosen": -0.25783437490463257, |
| "logits/rejected": -0.2909075915813446, |
| "logps/chosen": -9.45044994354248, |
| "logps/rejected": -33.906532287597656, |
| "loss": 0.8592720627784729, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4804239869117737, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.12236574292182922, |
| "rewards/margins": 1.7898168563842773, |
| "rewards/rejected": -1.9121828079223633, |
| "step": 152, |
| "train_speed(iter/s)": 0.010684 |
| }, |
| { |
| "epoch": 0.3414225941422594, |
| "grad_norm": 1.8399463891983032, |
| "learning_rate": 0.00019215672319306565, |
| "logits/chosen": -0.27110740542411804, |
| "logits/rejected": -0.3448094129562378, |
| "logps/chosen": -6.487441062927246, |
| "logps/rejected": -33.882781982421875, |
| "loss": 0.6851896047592163, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3750152587890625, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0861067920923233, |
| "rewards/margins": 2.013005018234253, |
| "rewards/rejected": -1.9268981218338013, |
| "step": 153, |
| "train_speed(iter/s)": 0.010684 |
| }, |
| { |
| "epoch": 0.34365411436541143, |
| "grad_norm": 1.2756677865982056, |
| "learning_rate": 0.00019201277882589487, |
| "logits/chosen": -0.23933258652687073, |
| "logits/rejected": -0.40502744913101196, |
| "logps/chosen": -4.149328231811523, |
| "logps/rejected": -56.27098083496094, |
| "loss": 0.3939424157142639, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.240768164396286, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.10984864085912704, |
| "rewards/margins": 3.799041271209717, |
| "rewards/rejected": -3.689192771911621, |
| "step": 154, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.3458856345885635, |
| "grad_norm": 2.6633708477020264, |
| "learning_rate": 0.00019186758048545498, |
| "logits/chosen": -0.2815844714641571, |
| "logits/rejected": -0.422237366437912, |
| "logps/chosen": -9.068172454833984, |
| "logps/rejected": -44.39888000488281, |
| "loss": 0.833120584487915, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5922810435295105, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": -0.021779753267765045, |
| "rewards/margins": 2.4418890476226807, |
| "rewards/rejected": -2.4636685848236084, |
| "step": 155, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.34811715481171546, |
| "grad_norm": 3.4869775772094727, |
| "learning_rate": 0.00019172113015054532, |
| "logits/chosen": -0.38594380021095276, |
| "logits/rejected": -0.4550650715827942, |
| "logps/chosen": -9.38473129272461, |
| "logps/rejected": -38.14994812011719, |
| "loss": 0.6951225996017456, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.44333207607269287, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07232007384300232, |
| "rewards/margins": 2.3049843311309814, |
| "rewards/rejected": -2.2326643466949463, |
| "step": 156, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.3503486750348675, |
| "grad_norm": 2.0695393085479736, |
| "learning_rate": 0.00019157342981702792, |
| "logits/chosen": -0.30112287402153015, |
| "logits/rejected": -0.4652690887451172, |
| "logps/chosen": -4.151251792907715, |
| "logps/rejected": -40.637821197509766, |
| "loss": 0.5484941005706787, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.34684789180755615, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.14294274151325226, |
| "rewards/margins": 2.3962206840515137, |
| "rewards/rejected": -2.2532777786254883, |
| "step": 157, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.35258019525801954, |
| "grad_norm": 2.7222163677215576, |
| "learning_rate": 0.0001914244814978001, |
| "logits/chosen": -0.3103567063808441, |
| "logits/rejected": -0.430208683013916, |
| "logps/chosen": -12.615564346313477, |
| "logps/rejected": -38.72933578491211, |
| "loss": 0.793499231338501, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5542211532592773, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.015316572040319443, |
| "rewards/margins": 2.3503148555755615, |
| "rewards/rejected": -2.334998369216919, |
| "step": 158, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.3548117154811716, |
| "grad_norm": 2.9450817108154297, |
| "learning_rate": 0.00019127428722276685, |
| "logits/chosen": -0.24224838614463806, |
| "logits/rejected": -0.43393394351005554, |
| "logps/chosen": -8.04431438446045, |
| "logps/rejected": -53.618770599365234, |
| "loss": 0.753520131111145, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5266157984733582, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.10486146807670593, |
| "rewards/margins": 3.499842643737793, |
| "rewards/rejected": -3.3949811458587646, |
| "step": 159, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.35704323570432356, |
| "grad_norm": 6.73776912689209, |
| "learning_rate": 0.0001911228490388136, |
| "logits/chosen": -0.3112480044364929, |
| "logits/rejected": -0.4515516757965088, |
| "logps/chosen": -9.544122695922852, |
| "logps/rejected": -44.96870803833008, |
| "loss": 0.7057361006736755, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5024272799491882, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.22010847926139832, |
| "rewards/margins": 2.9551522731781006, |
| "rewards/rejected": -2.735043525695801, |
| "step": 160, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.3592747559274756, |
| "grad_norm": 3.6044814586639404, |
| "learning_rate": 0.00019097016900977788, |
| "logits/chosen": -0.3657134771347046, |
| "logits/rejected": -0.45344415307044983, |
| "logps/chosen": -11.24877643585205, |
| "logps/rejected": -45.62288284301758, |
| "loss": 0.5986043810844421, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3538128137588501, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.06909701228141785, |
| "rewards/margins": 3.109391927719116, |
| "rewards/rejected": -3.040294885635376, |
| "step": 161, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.36150627615062764, |
| "grad_norm": 1.7556335926055908, |
| "learning_rate": 0.00019081624921642157, |
| "logits/chosen": -0.3330416977405548, |
| "logits/rejected": -0.49046823382377625, |
| "logps/chosen": -7.067329406738281, |
| "logps/rejected": -46.18006896972656, |
| "loss": 0.6361539363861084, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.39248889684677124, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.14040817320346832, |
| "rewards/margins": 2.906202793121338, |
| "rewards/rejected": -2.765794515609741, |
| "step": 162, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.3637377963737796, |
| "grad_norm": 1.4350221157073975, |
| "learning_rate": 0.00019066109175640224, |
| "logits/chosen": -0.38426709175109863, |
| "logits/rejected": -0.3983399271965027, |
| "logps/chosen": -12.697317123413086, |
| "logps/rejected": -50.39441680908203, |
| "loss": 0.7568346858024597, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4696040451526642, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.14664673805236816, |
| "rewards/margins": 3.4247310161590576, |
| "rewards/rejected": -3.2780845165252686, |
| "step": 163, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.36596931659693166, |
| "grad_norm": 2.5890817642211914, |
| "learning_rate": 0.00019050469874424476, |
| "logits/chosen": -0.32147228717803955, |
| "logits/rejected": -0.45530298352241516, |
| "logps/chosen": -8.81505012512207, |
| "logps/rejected": -47.758445739746094, |
| "loss": 0.8394909501075745, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5335902571678162, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.005741395056247711, |
| "rewards/margins": 3.229443073272705, |
| "rewards/rejected": -3.23518443107605, |
| "step": 164, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.3682008368200837, |
| "grad_norm": 2.527979850769043, |
| "learning_rate": 0.0001903470723113124, |
| "logits/chosen": -0.3687136173248291, |
| "logits/rejected": -0.36498233675956726, |
| "logps/chosen": -12.974212646484375, |
| "logps/rejected": -32.33938217163086, |
| "loss": 0.9435381889343262, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5042305588722229, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.09565389156341553, |
| "rewards/margins": 1.7492725849151611, |
| "rewards/rejected": -1.8449265956878662, |
| "step": 165, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.3704323570432357, |
| "grad_norm": 2.3836722373962402, |
| "learning_rate": 0.00019018821460577777, |
| "logits/chosen": -0.2881522476673126, |
| "logits/rejected": -0.39948752522468567, |
| "logps/chosen": -14.373290061950684, |
| "logps/rejected": -46.974891662597656, |
| "loss": 1.0135047435760498, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6972138285636902, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.46250468492507935, |
| "rewards/margins": 2.5046420097351074, |
| "rewards/rejected": -2.967146873474121, |
| "step": 166, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.3726638772663877, |
| "grad_norm": 2.324857234954834, |
| "learning_rate": 0.00019002812779259364, |
| "logits/chosen": -0.29876166582107544, |
| "logits/rejected": -0.47440457344055176, |
| "logps/chosen": -6.713703632354736, |
| "logps/rejected": -38.95991134643555, |
| "loss": 0.7639641761779785, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4946966767311096, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.18299029767513275, |
| "rewards/margins": 2.441457748413086, |
| "rewards/rejected": -2.258467435836792, |
| "step": 167, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.37489539748953976, |
| "grad_norm": 5.72042179107666, |
| "learning_rate": 0.00018986681405346322, |
| "logits/chosen": -0.28162381052970886, |
| "logits/rejected": -0.47184035181999207, |
| "logps/chosen": -9.439067840576172, |
| "logps/rejected": -53.3689079284668, |
| "loss": 0.964309811592102, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6164814233779907, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.01046261191368103, |
| "rewards/margins": 3.1293952465057373, |
| "rewards/rejected": -3.1398582458496094, |
| "step": 168, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.37712691771269174, |
| "grad_norm": 4.976809978485107, |
| "learning_rate": 0.00018970427558681082, |
| "logits/chosen": -0.25434356927871704, |
| "logits/rejected": -0.48102837800979614, |
| "logps/chosen": -8.21963119506836, |
| "logps/rejected": -65.61998748779297, |
| "loss": 0.8252819776535034, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.646252453327179, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1419672667980194, |
| "rewards/margins": 4.092907428741455, |
| "rewards/rejected": -3.9509406089782715, |
| "step": 169, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.3793584379358438, |
| "grad_norm": 3.3867485523223877, |
| "learning_rate": 0.0001895405146077514, |
| "logits/chosen": -0.34642744064331055, |
| "logits/rejected": -0.4897533357143402, |
| "logps/chosen": -5.762866020202637, |
| "logps/rejected": -46.86370086669922, |
| "loss": 0.7062004804611206, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.33989417552948, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.06110583245754242, |
| "rewards/margins": 2.4862968921661377, |
| "rewards/rejected": -2.5474026203155518, |
| "step": 170, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.3815899581589958, |
| "grad_norm": 1.8158555030822754, |
| "learning_rate": 0.00018937553334806077, |
| "logits/chosen": -0.2750059962272644, |
| "logits/rejected": -0.4136987030506134, |
| "logps/chosen": -8.728262901306152, |
| "logps/rejected": -45.89287185668945, |
| "loss": 0.7403351068496704, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.49930721521377563, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.16491073369979858, |
| "rewards/margins": 3.0132622718811035, |
| "rewards/rejected": -2.8483517169952393, |
| "step": 171, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.38382147838214786, |
| "grad_norm": 1.472412109375, |
| "learning_rate": 0.000189209334056145, |
| "logits/chosen": -0.27084222435951233, |
| "logits/rejected": -0.434164434671402, |
| "logps/chosen": -5.168161392211914, |
| "logps/rejected": -45.41401672363281, |
| "loss": 0.6193027496337891, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.42660266160964966, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.12104358524084091, |
| "rewards/margins": 2.716996431350708, |
| "rewards/rejected": -2.5959529876708984, |
| "step": 172, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.38605299860529985, |
| "grad_norm": 3.4075303077697754, |
| "learning_rate": 0.00018904191899700977, |
| "logits/chosen": -0.3655165433883667, |
| "logits/rejected": -0.41131308674812317, |
| "logps/chosen": -6.5695905685424805, |
| "logps/rejected": -32.802406311035156, |
| "loss": 0.7823173403739929, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4896646738052368, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.10015285015106201, |
| "rewards/margins": 1.8082492351531982, |
| "rewards/rejected": -1.7080962657928467, |
| "step": 173, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.3882845188284519, |
| "grad_norm": 1.7677769660949707, |
| "learning_rate": 0.0001888732904522296, |
| "logits/chosen": -0.3326292634010315, |
| "logits/rejected": -0.4164149761199951, |
| "logps/chosen": -5.277183532714844, |
| "logps/rejected": -42.28554916381836, |
| "loss": 0.6526207327842712, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4348045885562897, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.10670876502990723, |
| "rewards/margins": 2.7690269947052, |
| "rewards/rejected": -2.662318468093872, |
| "step": 174, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.3905160390516039, |
| "grad_norm": 2.253504753112793, |
| "learning_rate": 0.00018870345071991663, |
| "logits/chosen": -0.3795526325702667, |
| "logits/rejected": -0.4494696855545044, |
| "logps/chosen": -4.6198410987854, |
| "logps/rejected": -34.95786666870117, |
| "loss": 0.5223369598388672, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.29292282462120056, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.05781913921236992, |
| "rewards/margins": 2.233022928237915, |
| "rewards/rejected": -2.175203800201416, |
| "step": 175, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.3927475592747559, |
| "grad_norm": 4.090317726135254, |
| "learning_rate": 0.00018853240211468944, |
| "logits/chosen": -0.33655816316604614, |
| "logits/rejected": -0.3891099691390991, |
| "logps/chosen": -12.897551536560059, |
| "logps/rejected": -34.91562271118164, |
| "loss": 0.9967415928840637, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5328224301338196, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.0005232337862253189, |
| "rewards/margins": 1.7757031917572021, |
| "rewards/rejected": -1.7762264013290405, |
| "step": 176, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.39497907949790795, |
| "grad_norm": 1.5479252338409424, |
| "learning_rate": 0.0001883601469676414, |
| "logits/chosen": -0.31817540526390076, |
| "logits/rejected": -0.4586165249347687, |
| "logps/chosen": -4.228600025177002, |
| "logps/rejected": -44.335289001464844, |
| "loss": 0.5654776096343994, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.354885458946228, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.04608811438083649, |
| "rewards/margins": 2.883877754211426, |
| "rewards/rejected": -2.83778977394104, |
| "step": 177, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.39721059972106, |
| "grad_norm": 1.458658218383789, |
| "learning_rate": 0.00018818668762630888, |
| "logits/chosen": -0.3560311794281006, |
| "logits/rejected": -0.46038633584976196, |
| "logps/chosen": -6.351040363311768, |
| "logps/rejected": -38.50391387939453, |
| "loss": 0.5693007111549377, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3589138686656952, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.06642849743366241, |
| "rewards/margins": 2.5289833545684814, |
| "rewards/rejected": -2.462554693222046, |
| "step": 178, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.39944211994421197, |
| "grad_norm": 1.4788439273834229, |
| "learning_rate": 0.00018801202645463947, |
| "logits/chosen": -0.44740062952041626, |
| "logits/rejected": -0.48686325550079346, |
| "logps/chosen": -7.422541618347168, |
| "logps/rejected": -34.7234001159668, |
| "loss": 0.7467449903488159, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5351516008377075, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2607961595058441, |
| "rewards/margins": 2.635673999786377, |
| "rewards/rejected": -2.374877691268921, |
| "step": 179, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.401673640167364, |
| "grad_norm": 1.2784523963928223, |
| "learning_rate": 0.00018783616583295943, |
| "logits/chosen": -0.34507742524147034, |
| "logits/rejected": -0.46668770909309387, |
| "logps/chosen": -4.764893531799316, |
| "logps/rejected": -44.04815673828125, |
| "loss": 0.49975255131721497, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3011859357357025, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.152861088514328, |
| "rewards/margins": 2.924600124359131, |
| "rewards/rejected": -2.7717387676239014, |
| "step": 180, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.40390516039051605, |
| "grad_norm": 1.7386562824249268, |
| "learning_rate": 0.0001876591081579415, |
| "logits/chosen": -0.38848260045051575, |
| "logits/rejected": -0.48753347992897034, |
| "logps/chosen": -5.946084499359131, |
| "logps/rejected": -40.89453125, |
| "loss": 0.59507155418396, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.37911903858184814, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1187826544046402, |
| "rewards/margins": 2.863964319229126, |
| "rewards/rejected": -2.7451813220977783, |
| "step": 181, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.4061366806136681, |
| "grad_norm": 1.747938632965088, |
| "learning_rate": 0.0001874808558425722, |
| "logits/chosen": -0.3996775150299072, |
| "logits/rejected": -0.4858594536781311, |
| "logps/chosen": -8.173686027526855, |
| "logps/rejected": -36.68798065185547, |
| "loss": 0.7136989831924438, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4385717213153839, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.21724800765514374, |
| "rewards/margins": 2.21327543258667, |
| "rewards/rejected": -1.9960275888442993, |
| "step": 182, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.40836820083682007, |
| "grad_norm": 2.1921677589416504, |
| "learning_rate": 0.00018730141131611882, |
| "logits/chosen": -0.3768147826194763, |
| "logits/rejected": -0.4843496084213257, |
| "logps/chosen": -7.464463233947754, |
| "logps/rejected": -37.9464111328125, |
| "loss": 0.744111180305481, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.43925032019615173, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.062185730785131454, |
| "rewards/margins": 2.2055001258850098, |
| "rewards/rejected": -2.1433143615722656, |
| "step": 183, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.4105997210599721, |
| "grad_norm": 3.3944175243377686, |
| "learning_rate": 0.0001871207770240965, |
| "logits/chosen": -0.3745571970939636, |
| "logits/rejected": -0.5446013808250427, |
| "logps/chosen": -6.670903205871582, |
| "logps/rejected": -52.90716552734375, |
| "loss": 0.8081664443016052, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.546490490436554, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.06610552966594696, |
| "rewards/margins": 3.103334426879883, |
| "rewards/rejected": -3.037229061126709, |
| "step": 184, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.41283124128312415, |
| "grad_norm": 1.7184181213378906, |
| "learning_rate": 0.00018693895542823476, |
| "logits/chosen": -0.42309850454330444, |
| "logits/rejected": -0.5455373525619507, |
| "logps/chosen": -4.061583518981934, |
| "logps/rejected": -46.45550537109375, |
| "loss": 0.5212042331695557, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3185817003250122, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.06332439184188843, |
| "rewards/margins": 3.2307968139648438, |
| "rewards/rejected": -3.1674723625183105, |
| "step": 185, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.41506276150627613, |
| "grad_norm": 2.1810686588287354, |
| "learning_rate": 0.00018675594900644394, |
| "logits/chosen": -0.43507513403892517, |
| "logits/rejected": -0.5661131739616394, |
| "logps/chosen": -7.813204765319824, |
| "logps/rejected": -47.85923385620117, |
| "loss": 0.6977582573890686, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4816473722457886, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.05514959990978241, |
| "rewards/margins": 3.3984975814819336, |
| "rewards/rejected": -3.3433480262756348, |
| "step": 186, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.4172942817294282, |
| "grad_norm": 8.446174621582031, |
| "learning_rate": 0.0001865717602527816, |
| "logits/chosen": -0.3857502043247223, |
| "logits/rejected": -0.5707287788391113, |
| "logps/chosen": -5.875633239746094, |
| "logps/rejected": -52.29627990722656, |
| "loss": 0.8916729688644409, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6320705413818359, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.10125800967216492, |
| "rewards/margins": 3.7010817527770996, |
| "rewards/rejected": -3.599823474884033, |
| "step": 187, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.4195258019525802, |
| "grad_norm": 1.2875564098358154, |
| "learning_rate": 0.00018638639167741834, |
| "logits/chosen": -0.45664966106414795, |
| "logits/rejected": -0.6208757162094116, |
| "logps/chosen": -6.463584899902344, |
| "logps/rejected": -63.02524185180664, |
| "loss": 0.5230106115341187, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.38262802362442017, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.25487378239631653, |
| "rewards/margins": 4.994178295135498, |
| "rewards/rejected": -4.739304065704346, |
| "step": 188, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.4217573221757322, |
| "grad_norm": 2.2951433658599854, |
| "learning_rate": 0.00018619984580660367, |
| "logits/chosen": -0.4468421936035156, |
| "logits/rejected": -0.590311586856842, |
| "logps/chosen": -7.629635334014893, |
| "logps/rejected": -60.66893768310547, |
| "loss": 0.6314592361450195, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4014724791049957, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": -0.0873238816857338, |
| "rewards/margins": 4.23132848739624, |
| "rewards/rejected": -4.318652629852295, |
| "step": 189, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.42398884239888424, |
| "grad_norm": 3.432699680328369, |
| "learning_rate": 0.00018601212518263156, |
| "logits/chosen": -0.4163021147251129, |
| "logits/rejected": -0.5166957378387451, |
| "logps/chosen": -10.489439010620117, |
| "logps/rejected": -59.12989044189453, |
| "loss": 0.9744394421577454, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.7091968655586243, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.017084941267967224, |
| "rewards/margins": 4.347151279449463, |
| "rewards/rejected": -4.330066680908203, |
| "step": 190, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.4262203626220363, |
| "grad_norm": 3.6154932975769043, |
| "learning_rate": 0.00018582323236380588, |
| "logits/chosen": -0.5206790566444397, |
| "logits/rejected": -0.5340564250946045, |
| "logps/chosen": -14.040782928466797, |
| "logps/rejected": -46.77815628051758, |
| "loss": 1.0751701593399048, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.665191113948822, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.24230876564979553, |
| "rewards/margins": 3.153132438659668, |
| "rewards/rejected": -3.3954415321350098, |
| "step": 191, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.42845188284518826, |
| "grad_norm": 8.427726745605469, |
| "learning_rate": 0.00018563316992440543, |
| "logits/chosen": -0.4730026423931122, |
| "logits/rejected": -0.6191343069076538, |
| "logps/chosen": -6.7660651206970215, |
| "logps/rejected": -55.98576354980469, |
| "loss": 0.8084225058555603, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5363692045211792, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.10752980411052704, |
| "rewards/margins": 3.833192825317383, |
| "rewards/rejected": -3.940722703933716, |
| "step": 192, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.4306834030683403, |
| "grad_norm": 2.5356807708740234, |
| "learning_rate": 0.00018544194045464886, |
| "logits/chosen": -0.41117027401924133, |
| "logits/rejected": -0.5410706400871277, |
| "logps/chosen": -6.973834991455078, |
| "logps/rejected": -47.055511474609375, |
| "loss": 0.7811009883880615, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5517580509185791, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.13715188205242157, |
| "rewards/margins": 3.0469465255737305, |
| "rewards/rejected": -2.909794330596924, |
| "step": 193, |
| "train_speed(iter/s)": 0.010692 |
| }, |
| { |
| "epoch": 0.43291492329149234, |
| "grad_norm": 1.7070624828338623, |
| "learning_rate": 0.00018524954656065946, |
| "logits/chosen": -0.46515780687332153, |
| "logits/rejected": -0.5741381645202637, |
| "logps/chosen": -7.333560943603516, |
| "logps/rejected": -46.890380859375, |
| "loss": 0.46373608708381653, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3009029030799866, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.09283138811588287, |
| "rewards/margins": 3.2314822673797607, |
| "rewards/rejected": -3.138651132583618, |
| "step": 194, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.4351464435146444, |
| "grad_norm": 2.165066957473755, |
| "learning_rate": 0.00018505599086442956, |
| "logits/chosen": -0.4237922728061676, |
| "logits/rejected": -0.6133349537849426, |
| "logps/chosen": -2.157940626144409, |
| "logps/rejected": -44.79521942138672, |
| "loss": 0.48500534892082214, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.28576743602752686, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1598954200744629, |
| "rewards/margins": 2.9778366088867188, |
| "rewards/rejected": -2.817941188812256, |
| "step": 195, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.43737796373779636, |
| "grad_norm": 2.6087758541107178, |
| "learning_rate": 0.0001848612760037848, |
| "logits/chosen": -0.42983922362327576, |
| "logits/rejected": -0.5865707397460938, |
| "logps/chosen": -4.183155536651611, |
| "logps/rejected": -44.91338348388672, |
| "loss": 0.4380960762500763, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.22082072496414185, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1885657161474228, |
| "rewards/margins": 2.626762628555298, |
| "rewards/rejected": -2.438196897506714, |
| "step": 196, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.4396094839609484, |
| "grad_norm": 2.3417139053344727, |
| "learning_rate": 0.0001846654046323482, |
| "logits/chosen": -0.4529270827770233, |
| "logits/rejected": -0.5250231027603149, |
| "logps/chosen": -10.214118957519531, |
| "logps/rejected": -33.07242202758789, |
| "loss": 0.9449623823165894, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6152971982955933, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.03806466981768608, |
| "rewards/margins": 1.6490262746810913, |
| "rewards/rejected": -1.6870909929275513, |
| "step": 197, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.44184100418410044, |
| "grad_norm": 2.435980796813965, |
| "learning_rate": 0.0001844683794195041, |
| "logits/chosen": -0.41686204075813293, |
| "logits/rejected": -0.5872527956962585, |
| "logps/chosen": -5.5450968742370605, |
| "logps/rejected": -41.29132843017578, |
| "loss": 0.6443852782249451, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3510324954986572, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.10375159978866577, |
| "rewards/margins": 2.4684395790100098, |
| "rewards/rejected": -2.364687919616699, |
| "step": 198, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.4440725244072524, |
| "grad_norm": 2.279975414276123, |
| "learning_rate": 0.00018427020305036157, |
| "logits/chosen": -0.4846893548965454, |
| "logits/rejected": -0.5544926524162292, |
| "logps/chosen": -5.250210762023926, |
| "logps/rejected": -37.56634521484375, |
| "loss": 0.6850971579551697, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3965678811073303, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.009095368906855583, |
| "rewards/margins": 2.2770018577575684, |
| "rewards/rejected": -2.267906427383423, |
| "step": 199, |
| "train_speed(iter/s)": 0.010693 |
| }, |
| { |
| "epoch": 0.44630404463040446, |
| "grad_norm": 3.310929298400879, |
| "learning_rate": 0.00018407087822571794, |
| "logits/chosen": -0.45392054319381714, |
| "logits/rejected": -0.5642043352127075, |
| "logps/chosen": -9.782092094421387, |
| "logps/rejected": -40.09915542602539, |
| "loss": 0.8508836030960083, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6215177178382874, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1146484911441803, |
| "rewards/margins": 2.442267656326294, |
| "rewards/rejected": -2.3276190757751465, |
| "step": 200, |
| "train_speed(iter/s)": 0.010694 |
| }, |
| { |
| "epoch": 0.4485355648535565, |
| "grad_norm": 2.2623071670532227, |
| "learning_rate": 0.00018387040766202202, |
| "logits/chosen": -0.45955681800842285, |
| "logits/rejected": -0.5978164672851562, |
| "logps/chosen": -9.393783569335938, |
| "logps/rejected": -52.4150390625, |
| "loss": 0.6487902998924255, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4702387750148773, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.14986836910247803, |
| "rewards/margins": 3.318939208984375, |
| "rewards/rejected": -3.1690707206726074, |
| "step": 201, |
| "train_speed(iter/s)": 0.010683 |
| }, |
| { |
| "epoch": 0.4507670850767085, |
| "grad_norm": 1.4719666242599487, |
| "learning_rate": 0.00018366879409133702, |
| "logits/chosen": -0.5009268522262573, |
| "logits/rejected": -0.625489354133606, |
| "logps/chosen": -7.751692771911621, |
| "logps/rejected": -50.78370666503906, |
| "loss": 0.5859130620956421, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.39871278405189514, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.10605036467313766, |
| "rewards/margins": 3.2280044555664062, |
| "rewards/rejected": -3.1219542026519775, |
| "step": 202, |
| "train_speed(iter/s)": 0.010683 |
| }, |
| { |
| "epoch": 0.4529986052998605, |
| "grad_norm": 3.2725584506988525, |
| "learning_rate": 0.00018346604026130335, |
| "logits/chosen": -0.5253587961196899, |
| "logits/rejected": -0.6153156161308289, |
| "logps/chosen": -7.448980331420898, |
| "logps/rejected": -37.25505828857422, |
| "loss": 0.8574545383453369, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5657089352607727, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08424842357635498, |
| "rewards/margins": 2.4470019340515137, |
| "rewards/rejected": -2.3627536296844482, |
| "step": 203, |
| "train_speed(iter/s)": 0.010683 |
| }, |
| { |
| "epoch": 0.45523012552301256, |
| "grad_norm": 1.6742641925811768, |
| "learning_rate": 0.00018326214893510113, |
| "logits/chosen": -0.5298970937728882, |
| "logits/rejected": -0.5919500589370728, |
| "logps/chosen": -6.125566005706787, |
| "logps/rejected": -37.60496139526367, |
| "loss": 0.637344479560852, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3929066061973572, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.13108696043491364, |
| "rewards/margins": 2.768951654434204, |
| "rewards/rejected": -2.637864589691162, |
| "step": 204, |
| "train_speed(iter/s)": 0.010684 |
| }, |
| { |
| "epoch": 0.45746164574616455, |
| "grad_norm": 1.8297412395477295, |
| "learning_rate": 0.00018305712289141263, |
| "logits/chosen": -0.5556704998016357, |
| "logits/rejected": -0.6641337871551514, |
| "logps/chosen": -7.0842132568359375, |
| "logps/rejected": -43.63640213012695, |
| "loss": 0.611258327960968, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3571435213088989, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.08243419229984283, |
| "rewards/margins": 2.901592969894409, |
| "rewards/rejected": -2.8191587924957275, |
| "step": 205, |
| "train_speed(iter/s)": 0.010684 |
| }, |
| { |
| "epoch": 0.4596931659693166, |
| "grad_norm": 2.103322982788086, |
| "learning_rate": 0.00018285096492438424, |
| "logits/chosen": -0.5411264896392822, |
| "logits/rejected": -0.6788846254348755, |
| "logps/chosen": -7.095435619354248, |
| "logps/rejected": -55.614654541015625, |
| "loss": 0.5769423842430115, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3610311448574066, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.028350576758384705, |
| "rewards/margins": 3.771873950958252, |
| "rewards/rejected": -3.743523120880127, |
| "step": 206, |
| "train_speed(iter/s)": 0.010684 |
| }, |
| { |
| "epoch": 0.4619246861924686, |
| "grad_norm": 5.2644124031066895, |
| "learning_rate": 0.00018264367784358854, |
| "logits/chosen": -0.5137989521026611, |
| "logits/rejected": -0.6350421905517578, |
| "logps/chosen": -8.56877326965332, |
| "logps/rejected": -47.056922912597656, |
| "loss": 0.8458958268165588, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6215453743934631, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1831740289926529, |
| "rewards/margins": 3.3318536281585693, |
| "rewards/rejected": -3.148679494857788, |
| "step": 207, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.46415620641562066, |
| "grad_norm": 2.3040452003479004, |
| "learning_rate": 0.00018243526447398595, |
| "logits/chosen": -0.5189839005470276, |
| "logits/rejected": -0.6215395927429199, |
| "logps/chosen": -8.22704792022705, |
| "logps/rejected": -45.81852340698242, |
| "loss": 0.6660462617874146, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.42924296855926514, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.015878167003393173, |
| "rewards/margins": 3.395660877227783, |
| "rewards/rejected": -3.3797831535339355, |
| "step": 208, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.46638772663877265, |
| "grad_norm": 4.965414047241211, |
| "learning_rate": 0.00018222572765588625, |
| "logits/chosen": -0.4948608875274658, |
| "logits/rejected": -0.5670303106307983, |
| "logps/chosen": -14.887557983398438, |
| "logps/rejected": -48.515113830566406, |
| "loss": 0.8501808643341064, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5177828669548035, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.011799663305282593, |
| "rewards/margins": 3.1876440048217773, |
| "rewards/rejected": -3.175844192504883, |
| "step": 209, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4686192468619247, |
| "grad_norm": 2.125880479812622, |
| "learning_rate": 0.00018201507024490988, |
| "logits/chosen": -0.4645600914955139, |
| "logits/rejected": -0.5849769711494446, |
| "logps/chosen": -14.342391967773438, |
| "logps/rejected": -53.85160446166992, |
| "loss": 0.7610176205635071, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.60414057970047, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.1484633982181549, |
| "rewards/margins": 3.265408992767334, |
| "rewards/rejected": -3.11694598197937, |
| "step": 210, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4708507670850767, |
| "grad_norm": 1.3584539890289307, |
| "learning_rate": 0.0001818032951119489, |
| "logits/chosen": -0.4379713833332062, |
| "logits/rejected": -0.6403225660324097, |
| "logps/chosen": -3.4099626541137695, |
| "logps/rejected": -55.08962631225586, |
| "loss": 0.4291754364967346, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.28185123205184937, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2514503002166748, |
| "rewards/margins": 3.88580584526062, |
| "rewards/rejected": -3.6343557834625244, |
| "step": 211, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4730822873082287, |
| "grad_norm": 1.7748886346817017, |
| "learning_rate": 0.00018159040514312803, |
| "logits/chosen": -0.5069133043289185, |
| "logits/rejected": -0.5805174708366394, |
| "logps/chosen": -8.58336067199707, |
| "logps/rejected": -35.97822189331055, |
| "loss": 0.663266658782959, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3847760558128357, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2307383120059967, |
| "rewards/margins": 2.5746662616729736, |
| "rewards/rejected": -2.343928098678589, |
| "step": 212, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.47531380753138075, |
| "grad_norm": 1.843264102935791, |
| "learning_rate": 0.00018137640323976536, |
| "logits/chosen": -0.4634440243244171, |
| "logits/rejected": -0.5595231056213379, |
| "logps/chosen": -12.369077682495117, |
| "logps/rejected": -49.64701843261719, |
| "loss": 0.7064059972763062, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5452557802200317, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.06534570455551147, |
| "rewards/margins": 3.548327684402466, |
| "rewards/rejected": -3.4829816818237305, |
| "step": 213, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4775453277545328, |
| "grad_norm": 3.4433913230895996, |
| "learning_rate": 0.00018116129231833249, |
| "logits/chosen": -0.4008879065513611, |
| "logits/rejected": -0.5341296195983887, |
| "logps/chosen": -8.161286354064941, |
| "logps/rejected": -41.809959411621094, |
| "loss": 0.7387677431106567, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4233713746070862, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.04444505646824837, |
| "rewards/margins": 2.688593626022339, |
| "rewards/rejected": -2.644148588180542, |
| "step": 214, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4797768479776848, |
| "grad_norm": 2.6310174465179443, |
| "learning_rate": 0.00018094507531041517, |
| "logits/chosen": -0.4581325352191925, |
| "logits/rejected": -0.5395053625106812, |
| "logps/chosen": -12.050057411193848, |
| "logps/rejected": -46.67640686035156, |
| "loss": 0.6482417583465576, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4028882384300232, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.03946290165185928, |
| "rewards/margins": 3.062953233718872, |
| "rewards/rejected": -3.0234906673431396, |
| "step": 215, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4820083682008368, |
| "grad_norm": 2.039609432220459, |
| "learning_rate": 0.00018072775516267306, |
| "logits/chosen": -0.40940549969673157, |
| "logits/rejected": -0.5348517894744873, |
| "logps/chosen": -5.345945835113525, |
| "logps/rejected": -55.2205696105957, |
| "loss": 0.588547945022583, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.35309699177742004, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.013576401397585869, |
| "rewards/margins": 3.3509364128112793, |
| "rewards/rejected": -3.33735990524292, |
| "step": 216, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.48423988842398885, |
| "grad_norm": 1.844130039215088, |
| "learning_rate": 0.00018050933483679976, |
| "logits/chosen": -0.388172447681427, |
| "logits/rejected": -0.5932977199554443, |
| "logps/chosen": -7.16074800491333, |
| "logps/rejected": -57.54563903808594, |
| "loss": 0.5982820987701416, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3665384352207184, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12162206321954727, |
| "rewards/margins": 3.5915048122406006, |
| "rewards/rejected": -3.4698829650878906, |
| "step": 217, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4864714086471409, |
| "grad_norm": 3.176985263824463, |
| "learning_rate": 0.0001802898173094824, |
| "logits/chosen": -0.4476504623889923, |
| "logits/rejected": -0.6101452112197876, |
| "logps/chosen": -8.568498611450195, |
| "logps/rejected": -55.76399612426758, |
| "loss": 0.759227454662323, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5219362378120422, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.05749676004052162, |
| "rewards/margins": 4.200704574584961, |
| "rewards/rejected": -4.143207550048828, |
| "step": 218, |
| "train_speed(iter/s)": 0.010685 |
| }, |
| { |
| "epoch": 0.4887029288702929, |
| "grad_norm": 8.979230880737305, |
| "learning_rate": 0.0001800692055723609, |
| "logits/chosen": -0.5705654621124268, |
| "logits/rejected": -0.6717063188552856, |
| "logps/chosen": -9.38418197631836, |
| "logps/rejected": -48.490394592285156, |
| "loss": 0.9973833560943604, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.765608549118042, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.36961740255355835, |
| "rewards/margins": 3.4103264808654785, |
| "rewards/rejected": -3.7799437046051025, |
| "step": 219, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.4909344490934449, |
| "grad_norm": 13.02912425994873, |
| "learning_rate": 0.0001798475026319875, |
| "logits/chosen": -0.48003196716308594, |
| "logits/rejected": -0.6781996488571167, |
| "logps/chosen": -7.269015789031982, |
| "logps/rejected": -62.30265808105469, |
| "loss": 0.7909015417098999, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.53118896484375, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": -0.15388834476470947, |
| "rewards/margins": 4.509274959564209, |
| "rewards/rejected": -4.663163185119629, |
| "step": 220, |
| "train_speed(iter/s)": 0.010686 |
| }, |
| { |
| "epoch": 0.49316596931659695, |
| "grad_norm": 6.664190292358398, |
| "learning_rate": 0.00017962471150978563, |
| "logits/chosen": -0.5324437022209167, |
| "logits/rejected": -0.6654635667800903, |
| "logps/chosen": -7.194028377532959, |
| "logps/rejected": -59.914695739746094, |
| "loss": 0.8783823251724243, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6288973689079285, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.018478773534297943, |
| "rewards/margins": 4.640980243682861, |
| "rewards/rejected": -4.622502326965332, |
| "step": 221, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.49539748953974894, |
| "grad_norm": 2.010087251663208, |
| "learning_rate": 0.00017940083524200858, |
| "logits/chosen": -0.43658989667892456, |
| "logits/rejected": -0.5585737228393555, |
| "logps/chosen": -8.627237319946289, |
| "logps/rejected": -48.756126403808594, |
| "loss": 0.5573357939720154, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.39346298575401306, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.34565243124961853, |
| "rewards/margins": 3.3448474407196045, |
| "rewards/rejected": -2.999195098876953, |
| "step": 222, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.497629009762901, |
| "grad_norm": 2.036358118057251, |
| "learning_rate": 0.00017917587687969847, |
| "logits/chosen": -0.434632807970047, |
| "logits/rejected": -0.5482152104377747, |
| "logps/chosen": -11.499567985534668, |
| "logps/rejected": -41.673377990722656, |
| "loss": 0.8241864442825317, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6470968723297119, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.23597590625286102, |
| "rewards/margins": 2.9532856941223145, |
| "rewards/rejected": -2.7173097133636475, |
| "step": 223, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.499860529986053, |
| "grad_norm": 2.800816297531128, |
| "learning_rate": 0.00017894983948864431, |
| "logits/chosen": -0.42586764693260193, |
| "logits/rejected": -0.5355687141418457, |
| "logps/chosen": -8.567970275878906, |
| "logps/rejected": -38.82350158691406, |
| "loss": 0.8209130167961121, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5143148899078369, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.0072904713451862335, |
| "rewards/margins": 2.314070224761963, |
| "rewards/rejected": -2.3213610649108887, |
| "step": 224, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.502092050209205, |
| "grad_norm": 3.365342140197754, |
| "learning_rate": 0.0001787227261493405, |
| "logits/chosen": -0.423734188079834, |
| "logits/rejected": -0.568479061126709, |
| "logps/chosen": -7.375796794891357, |
| "logps/rejected": -49.31288146972656, |
| "loss": 0.7071059942245483, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.46694087982177734, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.004096744582056999, |
| "rewards/margins": 3.2242605686187744, |
| "rewards/rejected": -3.2283570766448975, |
| "step": 225, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.5043235704323571, |
| "grad_norm": 1.2737809419631958, |
| "learning_rate": 0.0001784945399569447, |
| "logits/chosen": -0.36446237564086914, |
| "logits/rejected": -0.5933336615562439, |
| "logps/chosen": -3.509737730026245, |
| "logps/rejected": -58.02750778198242, |
| "loss": 0.33625951409339905, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.22468024492263794, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.18253092467784882, |
| "rewards/margins": 3.6850192546844482, |
| "rewards/rejected": -3.502488136291504, |
| "step": 226, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.5065550906555091, |
| "grad_norm": 2.068448305130005, |
| "learning_rate": 0.00017826528402123564, |
| "logits/chosen": -0.42301344871520996, |
| "logits/rejected": -0.5985809564590454, |
| "logps/chosen": -2.5439209938049316, |
| "logps/rejected": -45.29689025878906, |
| "loss": 0.36974459886550903, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.231589674949646, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2267550826072693, |
| "rewards/margins": 3.147808790206909, |
| "rewards/rejected": -2.9210541248321533, |
| "step": 227, |
| "train_speed(iter/s)": 0.010687 |
| }, |
| { |
| "epoch": 0.5087866108786611, |
| "grad_norm": 1.3184460401535034, |
| "learning_rate": 0.00017803496146657084, |
| "logits/chosen": -0.4008505344390869, |
| "logits/rejected": -0.592756450176239, |
| "logps/chosen": -6.708431243896484, |
| "logps/rejected": -48.843502044677734, |
| "loss": 0.6300328969955444, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.46876925230026245, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.32074451446533203, |
| "rewards/margins": 3.1505868434906006, |
| "rewards/rejected": -2.8298423290252686, |
| "step": 228, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.5110181311018132, |
| "grad_norm": 2.306436538696289, |
| "learning_rate": 0.00017780357543184397, |
| "logits/chosen": -0.4543289244174957, |
| "logits/rejected": -0.5702416300773621, |
| "logps/chosen": -7.456311225891113, |
| "logps/rejected": -55.43143081665039, |
| "loss": 0.6797499060630798, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.43957531452178955, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.061677221208810806, |
| "rewards/margins": 4.144193649291992, |
| "rewards/rejected": -4.082516670227051, |
| "step": 229, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.5132496513249651, |
| "grad_norm": 5.570302963256836, |
| "learning_rate": 0.000177571129070442, |
| "logits/chosen": -0.5052129626274109, |
| "logits/rejected": -0.5332801342010498, |
| "logps/chosen": -10.987539291381836, |
| "logps/rejected": -43.629920959472656, |
| "loss": 0.8144917488098145, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5699902772903442, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.07897716760635376, |
| "rewards/margins": 2.79160213470459, |
| "rewards/rejected": -2.870579481124878, |
| "step": 230, |
| "train_speed(iter/s)": 0.010688 |
| }, |
| { |
| "epoch": 0.5154811715481171, |
| "grad_norm": 1.758008360862732, |
| "learning_rate": 0.00017733762555020234, |
| "logits/chosen": -0.49660688638687134, |
| "logits/rejected": -0.5793140530586243, |
| "logps/chosen": -7.395486354827881, |
| "logps/rejected": -45.230892181396484, |
| "loss": 0.4777700901031494, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3038936257362366, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.13568957149982452, |
| "rewards/margins": 3.222832202911377, |
| "rewards/rejected": -3.0871424674987793, |
| "step": 231, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5177126917712692, |
| "grad_norm": 1.362999439239502, |
| "learning_rate": 0.0001771030680533697, |
| "logits/chosen": -0.4924534261226654, |
| "logits/rejected": -0.6207826733589172, |
| "logps/chosen": -4.7806715965271, |
| "logps/rejected": -48.792823791503906, |
| "loss": 0.4094929099082947, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.2229592204093933, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2055450677871704, |
| "rewards/margins": 3.34159779548645, |
| "rewards/rejected": -3.1360526084899902, |
| "step": 232, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5199442119944212, |
| "grad_norm": 1.446892261505127, |
| "learning_rate": 0.00017686745977655254, |
| "logits/chosen": -0.5510218739509583, |
| "logits/rejected": -0.5702377557754517, |
| "logps/chosen": -7.457989692687988, |
| "logps/rejected": -40.870391845703125, |
| "loss": 0.6166619658470154, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.37686091661453247, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.019258877262473106, |
| "rewards/margins": 2.9649226665496826, |
| "rewards/rejected": -2.9456636905670166, |
| "step": 233, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5221757322175732, |
| "grad_norm": 2.5673022270202637, |
| "learning_rate": 0.00017663080393067967, |
| "logits/chosen": -0.4004722833633423, |
| "logits/rejected": -0.6144905090332031, |
| "logps/chosen": -8.209660530090332, |
| "logps/rejected": -64.06336975097656, |
| "loss": 0.4909789562225342, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.31386396288871765, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.20118242502212524, |
| "rewards/margins": 4.270333290100098, |
| "rewards/rejected": -4.069150447845459, |
| "step": 234, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5244072524407253, |
| "grad_norm": 4.679993152618408, |
| "learning_rate": 0.00017639310374095642, |
| "logits/chosen": -0.47914889454841614, |
| "logits/rejected": -0.6579112410545349, |
| "logps/chosen": -8.027458190917969, |
| "logps/rejected": -49.39167022705078, |
| "loss": 0.6046391129493713, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4407365620136261, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.19219349324703217, |
| "rewards/margins": 3.430323362350464, |
| "rewards/rejected": -3.2381296157836914, |
| "step": 235, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5266387726638773, |
| "grad_norm": 2.3275198936462402, |
| "learning_rate": 0.00017615436244682067, |
| "logits/chosen": -0.48543357849121094, |
| "logits/rejected": -0.6728047132492065, |
| "logps/chosen": -5.8822832107543945, |
| "logps/rejected": -53.12212371826172, |
| "loss": 0.7272516489028931, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5779345035552979, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.11761157214641571, |
| "rewards/margins": 3.723693370819092, |
| "rewards/rejected": -3.606081962585449, |
| "step": 236, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5288702928870292, |
| "grad_norm": 1.9778822660446167, |
| "learning_rate": 0.0001759145833018988, |
| "logits/chosen": -0.5121138095855713, |
| "logits/rejected": -0.581526517868042, |
| "logps/chosen": -11.055684089660645, |
| "logps/rejected": -48.428466796875, |
| "loss": 0.6768224835395813, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4439318776130676, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07668354362249374, |
| "rewards/margins": 3.0230700969696045, |
| "rewards/rejected": -2.9463863372802734, |
| "step": 237, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5311018131101813, |
| "grad_norm": 2.473088264465332, |
| "learning_rate": 0.00017567376957396125, |
| "logits/chosen": -0.4359075427055359, |
| "logits/rejected": -0.6152986288070679, |
| "logps/chosen": -8.427140235900879, |
| "logps/rejected": -43.27656555175781, |
| "loss": 0.8277904391288757, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5778707265853882, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2914345860481262, |
| "rewards/margins": 2.8861725330352783, |
| "rewards/rejected": -2.594738006591797, |
| "step": 238, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5333333333333333, |
| "grad_norm": 4.2344560623168945, |
| "learning_rate": 0.00017543192454487795, |
| "logits/chosen": -0.5183901786804199, |
| "logits/rejected": -0.648127555847168, |
| "logps/chosen": -7.237810134887695, |
| "logps/rejected": -46.60614776611328, |
| "loss": 0.5062097311019897, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3680287003517151, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.09910139441490173, |
| "rewards/margins": 3.3295600414276123, |
| "rewards/rejected": -3.2304584980010986, |
| "step": 239, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.5355648535564853, |
| "grad_norm": 12.738329887390137, |
| "learning_rate": 0.0001751890515105738, |
| "logits/chosen": -0.5236122608184814, |
| "logits/rejected": -0.6659674644470215, |
| "logps/chosen": -6.250277042388916, |
| "logps/rejected": -48.39547348022461, |
| "loss": 0.8066755533218384, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.5374718904495239, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.011939210817217827, |
| "rewards/margins": 3.1737723350524902, |
| "rewards/rejected": -3.1618332862854004, |
| "step": 240, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.5377963737796374, |
| "grad_norm": 1.7081589698791504, |
| "learning_rate": 0.00017494515378098347, |
| "logits/chosen": -0.49652621150016785, |
| "logits/rejected": -0.6365297436714172, |
| "logps/chosen": -4.525838375091553, |
| "logps/rejected": -44.274539947509766, |
| "loss": 0.4856886863708496, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.26593509316444397, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.10261054337024689, |
| "rewards/margins": 3.1593854427337646, |
| "rewards/rejected": -3.056774616241455, |
| "step": 241, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.5400278940027894, |
| "grad_norm": 1.5723114013671875, |
| "learning_rate": 0.00017470023468000654, |
| "logits/chosen": -0.5223122835159302, |
| "logits/rejected": -0.6690611839294434, |
| "logps/chosen": -3.2274744510650635, |
| "logps/rejected": -53.48045349121094, |
| "loss": 0.3347703516483307, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.19753773510456085, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1409139484167099, |
| "rewards/margins": 3.965757369995117, |
| "rewards/rejected": -3.824843406677246, |
| "step": 242, |
| "train_speed(iter/s)": 0.010691 |
| }, |
| { |
| "epoch": 0.5422594142259414, |
| "grad_norm": 1.1658532619476318, |
| "learning_rate": 0.00017445429754546209, |
| "logits/chosen": -0.4327467083930969, |
| "logits/rejected": -0.5850556492805481, |
| "logps/chosen": -6.455155372619629, |
| "logps/rejected": -53.67913055419922, |
| "loss": 0.45557722449302673, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3267042934894562, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.24175924062728882, |
| "rewards/margins": 3.549290418624878, |
| "rewards/rejected": -3.3075313568115234, |
| "step": 243, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.5444909344490935, |
| "grad_norm": 1.9491972923278809, |
| "learning_rate": 0.00017420734572904309, |
| "logits/chosen": -0.4307960569858551, |
| "logits/rejected": -0.6243136525154114, |
| "logps/chosen": -5.606278419494629, |
| "logps/rejected": -51.735652923583984, |
| "loss": 0.5139089226722717, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3616556227207184, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1640806794166565, |
| "rewards/margins": 3.619340419769287, |
| "rewards/rejected": -3.4552597999572754, |
| "step": 244, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.5467224546722455, |
| "grad_norm": 2.2012929916381836, |
| "learning_rate": 0.00017395938259627102, |
| "logits/chosen": -0.343536376953125, |
| "logits/rejected": -0.569451630115509, |
| "logps/chosen": -5.0532708168029785, |
| "logps/rejected": -62.23390197753906, |
| "loss": 0.3923553228378296, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.29478776454925537, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.23885516822338104, |
| "rewards/margins": 4.132390975952148, |
| "rewards/rejected": -3.893535852432251, |
| "step": 245, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.5489539748953974, |
| "grad_norm": 3.378282308578491, |
| "learning_rate": 0.00017371041152644976, |
| "logits/chosen": -0.4609750807285309, |
| "logits/rejected": -0.6062232851982117, |
| "logps/chosen": -6.916518211364746, |
| "logps/rejected": -52.16080856323242, |
| "loss": 0.6823198199272156, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.45661455392837524, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.3001694083213806, |
| "rewards/margins": 3.915588855743408, |
| "rewards/rejected": -3.6154191493988037, |
| "step": 246, |
| "train_speed(iter/s)": 0.01069 |
| }, |
| { |
| "epoch": 0.5511854951185495, |
| "grad_norm": 2.7868404388427734, |
| "learning_rate": 0.00017346043591261957, |
| "logits/chosen": -0.4002017676830292, |
| "logits/rejected": -0.6774240732192993, |
| "logps/chosen": -7.5064496994018555, |
| "logps/rejected": -64.324951171875, |
| "loss": 0.5472509860992432, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4029460847377777, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3002982437610626, |
| "rewards/margins": 4.144438743591309, |
| "rewards/rejected": -3.8441405296325684, |
| "step": 247, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5534170153417015, |
| "grad_norm": 2.1542208194732666, |
| "learning_rate": 0.0001732094591615109, |
| "logits/chosen": -0.34761539101600647, |
| "logits/rejected": -0.6237728595733643, |
| "logps/chosen": -8.17116641998291, |
| "logps/rejected": -72.38832092285156, |
| "loss": 0.5291301012039185, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.3634389340877533, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2887139618396759, |
| "rewards/margins": 4.893632888793945, |
| "rewards/rejected": -4.604918956756592, |
| "step": 248, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5556485355648536, |
| "grad_norm": 5.7039337158203125, |
| "learning_rate": 0.00017295748469349804, |
| "logits/chosen": -0.44048506021499634, |
| "logits/rejected": -0.7222636342048645, |
| "logps/chosen": -7.497531890869141, |
| "logps/rejected": -82.19099426269531, |
| "loss": 0.7382796406745911, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.6829304695129395, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.09070631116628647, |
| "rewards/margins": 6.177928924560547, |
| "rewards/rejected": -6.087222099304199, |
| "step": 249, |
| "train_speed(iter/s)": 0.010689 |
| }, |
| { |
| "epoch": 0.5578800557880056, |
| "grad_norm": 9.315430641174316, |
| "learning_rate": 0.00017270451594255233, |
| "logits/chosen": -0.5497281551361084, |
| "logits/rejected": -0.6751365065574646, |
| "logps/chosen": -8.304182052612305, |
| "logps/rejected": -60.03815460205078, |
| "loss": 0.5764181017875671, |
| "memory(GiB)": 46.81, |
| "nll_loss": 0.4564261734485626, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": -0.06676550954580307, |
| "rewards/margins": 4.525730133056641, |
| "rewards/rejected": -4.592494964599609, |
| "step": 250, |
| "train_speed(iter/s)": 0.01069 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 896, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.379183154429952e+17, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
