| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.35882484862076697, |
| "eval_steps": 300, |
| "global_step": 150, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.00239216565747178, |
| "grad_norm": 13.569995880126953, |
| "learning_rate": 3.1746031746031746e-06, |
| "logits/chosen": -0.6432992815971375, |
| "logits/rejected": -0.6486618518829346, |
| "logps/chosen": -10.907301902770996, |
| "logps/rejected": -27.245113372802734, |
| "loss": 1.6750074625015259, |
| "memory(GiB)": 45.5, |
| "nll_loss": 0.9818602204322815, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1, |
| "train_speed(iter/s)": 0.005543 |
| }, |
| { |
| "epoch": 0.00478433131494356, |
| "grad_norm": 6.833891868591309, |
| "learning_rate": 6.349206349206349e-06, |
| "logits/chosen": -0.6526631116867065, |
| "logits/rejected": -0.6589193344116211, |
| "logps/chosen": -13.911490440368652, |
| "logps/rejected": -26.267818450927734, |
| "loss": 1.5071839094161987, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.8140366077423096, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 2, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 0.00717649697241534, |
| "grad_norm": 4.8643999099731445, |
| "learning_rate": 9.523809523809523e-06, |
| "logits/chosen": -0.6420565247535706, |
| "logits/rejected": -0.6463347673416138, |
| "logps/chosen": -10.297296524047852, |
| "logps/rejected": -21.898191452026367, |
| "loss": 1.3684215545654297, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.6745598316192627, |
| "rewards/accuracies": 0.40625, |
| "rewards/chosen": 0.000917938828933984, |
| "rewards/margins": -0.0013222515117377043, |
| "rewards/rejected": 0.0022401907481253147, |
| "step": 3, |
| "train_speed(iter/s)": 0.005679 |
| }, |
| { |
| "epoch": 0.00956866262988712, |
| "grad_norm": 13.630061149597168, |
| "learning_rate": 1.2698412698412699e-05, |
| "logits/chosen": -0.6085749864578247, |
| "logits/rejected": -0.6145466566085815, |
| "logps/chosen": -9.835967063903809, |
| "logps/rejected": -20.03702163696289, |
| "loss": 1.6158230304718018, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.9227687120437622, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.002711558947339654, |
| "rewards/margins": 0.0003595724410843104, |
| "rewards/rejected": 0.0023519867099821568, |
| "step": 4, |
| "train_speed(iter/s)": 0.005709 |
| }, |
| { |
| "epoch": 0.0119608282873589, |
| "grad_norm": 9.080723762512207, |
| "learning_rate": 1.5873015873015872e-05, |
| "logits/chosen": -0.6302319765090942, |
| "logits/rejected": -0.6327166557312012, |
| "logps/chosen": -12.458501815795898, |
| "logps/rejected": -20.92184066772461, |
| "loss": 1.7525994777679443, |
| "memory(GiB)": 45.7, |
| "nll_loss": 1.0644183158874512, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.015323596075177193, |
| "rewards/margins": 0.010122949257493019, |
| "rewards/rejected": 0.005200647283345461, |
| "step": 5, |
| "train_speed(iter/s)": 0.005683 |
| }, |
| { |
| "epoch": 0.01435299394483068, |
| "grad_norm": 10.974760055541992, |
| "learning_rate": 1.9047619047619046e-05, |
| "logits/chosen": -0.6460072994232178, |
| "logits/rejected": -0.6503366231918335, |
| "logps/chosen": -8.966805458068848, |
| "logps/rejected": -29.11065101623535, |
| "loss": 1.2482141256332397, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.5670613050460815, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.03081551007926464, |
| "rewards/margins": 0.024802545085549355, |
| "rewards/rejected": 0.006012964062392712, |
| "step": 6, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 0.01674515960230246, |
| "grad_norm": 9.0870943069458, |
| "learning_rate": 2.2222222222222223e-05, |
| "logits/chosen": -0.5985114574432373, |
| "logits/rejected": -0.6029279232025146, |
| "logps/chosen": -12.831599235534668, |
| "logps/rejected": -22.661216735839844, |
| "loss": 1.4283000230789185, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.7556005120277405, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.05840035900473595, |
| "rewards/margins": 0.044253990054130554, |
| "rewards/rejected": 0.01414636429399252, |
| "step": 7, |
| "train_speed(iter/s)": 0.005641 |
| }, |
| { |
| "epoch": 0.01913732525977424, |
| "grad_norm": 7.071105003356934, |
| "learning_rate": 2.5396825396825397e-05, |
| "logits/chosen": -0.6822581887245178, |
| "logits/rejected": -0.6836903095245361, |
| "logps/chosen": -14.09546184539795, |
| "logps/rejected": -17.29296112060547, |
| "loss": 1.3682955503463745, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.6906765699386597, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.11469551920890808, |
| "rewards/margins": 0.042920827865600586, |
| "rewards/rejected": 0.07177469879388809, |
| "step": 8, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.02152949091724602, |
| "grad_norm": 6.927362442016602, |
| "learning_rate": 2.857142857142857e-05, |
| "logits/chosen": -0.6293038725852966, |
| "logits/rejected": -0.632675051689148, |
| "logps/chosen": -13.061110496520996, |
| "logps/rejected": -22.598844528198242, |
| "loss": 1.382551670074463, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.7284406423568726, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.11742030084133148, |
| "rewards/margins": 0.10634221881628036, |
| "rewards/rejected": 0.011078080162405968, |
| "step": 9, |
| "train_speed(iter/s)": 0.005614 |
| }, |
| { |
| "epoch": 0.0239216565747178, |
| "grad_norm": 3.728199005126953, |
| "learning_rate": 3.1746031746031745e-05, |
| "logits/chosen": -0.6618058681488037, |
| "logits/rejected": -0.6639875173568726, |
| "logps/chosen": -13.987442970275879, |
| "logps/rejected": -25.95743179321289, |
| "loss": 1.0269697904586792, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.4300592243671417, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.24149325489997864, |
| "rewards/margins": 0.27823469042778015, |
| "rewards/rejected": -0.03674142807722092, |
| "step": 10, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.02631382223218958, |
| "grad_norm": 7.327150344848633, |
| "learning_rate": 3.492063492063492e-05, |
| "logits/chosen": -0.6170505285263062, |
| "logits/rejected": -0.6200059056282043, |
| "logps/chosen": -12.316822052001953, |
| "logps/rejected": -22.407135009765625, |
| "loss": 1.2084441184997559, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.525244951248169, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.011396417394280434, |
| "rewards/margins": 0.15045319497585297, |
| "rewards/rejected": -0.13905677199363708, |
| "step": 11, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.02870598788966136, |
| "grad_norm": 5.15468168258667, |
| "learning_rate": 3.809523809523809e-05, |
| "logits/chosen": -0.6354009509086609, |
| "logits/rejected": -0.6367737054824829, |
| "logps/chosen": -16.074636459350586, |
| "logps/rejected": -23.24802017211914, |
| "loss": 1.352920413017273, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.6531318426132202, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": -0.10709743946790695, |
| "rewards/margins": 0.1539226770401001, |
| "rewards/rejected": -0.26102012395858765, |
| "step": 12, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.031098153547133137, |
| "grad_norm": 9.377359390258789, |
| "learning_rate": 4.126984126984127e-05, |
| "logits/chosen": -0.5982232093811035, |
| "logits/rejected": -0.5973206758499146, |
| "logps/chosen": -17.6663761138916, |
| "logps/rejected": -19.997753143310547, |
| "loss": 1.681210994720459, |
| "memory(GiB)": 45.7, |
| "nll_loss": 0.8636733889579773, |
| "rewards/accuracies": 0.40625, |
| "rewards/chosen": -0.24621205031871796, |
| "rewards/margins": -0.07156651467084885, |
| "rewards/rejected": -0.17464551329612732, |
| "step": 13, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.03349031920460492, |
| "grad_norm": 3.9681458473205566, |
| "learning_rate": 4.4444444444444447e-05, |
| "logits/chosen": -0.6217225193977356, |
| "logits/rejected": -0.6225230097770691, |
| "logps/chosen": -12.139883995056152, |
| "logps/rejected": -25.268104553222656, |
| "loss": 1.2384405136108398, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5618223547935486, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.0858367532491684, |
| "rewards/margins": 0.20134860277175903, |
| "rewards/rejected": -0.28718534111976624, |
| "step": 14, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.0358824848620767, |
| "grad_norm": 3.7261455059051514, |
| "learning_rate": 4.761904761904762e-05, |
| "logits/chosen": -0.5990071892738342, |
| "logits/rejected": -0.603517472743988, |
| "logps/chosen": -9.707769393920898, |
| "logps/rejected": -30.03302574157715, |
| "loss": 1.114662766456604, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5713129639625549, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.08963251113891602, |
| "rewards/margins": 0.5134348273277283, |
| "rewards/rejected": -0.42380231618881226, |
| "step": 15, |
| "train_speed(iter/s)": 0.005614 |
| }, |
| { |
| "epoch": 0.03827465051954848, |
| "grad_norm": 4.653024673461914, |
| "learning_rate": 5.0793650793650794e-05, |
| "logits/chosen": -0.6132736206054688, |
| "logits/rejected": -0.6161543726921082, |
| "logps/chosen": -15.514596939086914, |
| "logps/rejected": -26.068016052246094, |
| "loss": 1.281750202178955, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5498376488685608, |
| "rewards/accuracies": 0.4375, |
| "rewards/chosen": -0.09057736396789551, |
| "rewards/margins": 0.08230985701084137, |
| "rewards/rejected": -0.17288722097873688, |
| "step": 16, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.04066681617702026, |
| "grad_norm": 3.9547042846679688, |
| "learning_rate": 5.396825396825397e-05, |
| "logits/chosen": -0.5502676367759705, |
| "logits/rejected": -0.5536152124404907, |
| "logps/chosen": -13.769225120544434, |
| "logps/rejected": -27.007709503173828, |
| "loss": 1.413275122642517, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.7366797924041748, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": -0.05718833580613136, |
| "rewards/margins": 0.1863359659910202, |
| "rewards/rejected": -0.24352429807186127, |
| "step": 17, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.04305898183449204, |
| "grad_norm": 3.8449220657348633, |
| "learning_rate": 5.714285714285714e-05, |
| "logits/chosen": -0.5983889102935791, |
| "logits/rejected": -0.601371705532074, |
| "logps/chosen": -11.734331130981445, |
| "logps/rejected": -22.929622650146484, |
| "loss": 1.3582308292388916, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.639457106590271, |
| "rewards/accuracies": 0.4375, |
| "rewards/chosen": 0.03578497841954231, |
| "rewards/margins": 0.019625335931777954, |
| "rewards/rejected": 0.016159646213054657, |
| "step": 18, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 0.04545114749196382, |
| "grad_norm": 3.037893295288086, |
| "learning_rate": 6.0317460317460316e-05, |
| "logits/chosen": -0.6029090881347656, |
| "logits/rejected": -0.6060330867767334, |
| "logps/chosen": -13.030549049377441, |
| "logps/rejected": -24.52211570739746, |
| "loss": 1.2112592458724976, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.526079535484314, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.08603789657354355, |
| "rewards/margins": 0.11102047562599182, |
| "rewards/rejected": -0.024982573464512825, |
| "step": 19, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 0.0478433131494356, |
| "grad_norm": 3.4577159881591797, |
| "learning_rate": 6.349206349206349e-05, |
| "logits/chosen": -0.6348394155502319, |
| "logits/rejected": -0.6350866556167603, |
| "logps/chosen": -10.457337379455566, |
| "logps/rejected": -21.804182052612305, |
| "loss": 1.271092414855957, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5564278364181519, |
| "rewards/accuracies": 0.5, |
| "rewards/chosen": 0.13572227954864502, |
| "rewards/margins": 0.03713224455714226, |
| "rewards/rejected": 0.09859003126621246, |
| "step": 20, |
| "train_speed(iter/s)": 0.00565 |
| }, |
| { |
| "epoch": 0.05023547880690738, |
| "grad_norm": 2.758153200149536, |
| "learning_rate": 6.666666666666667e-05, |
| "logits/chosen": -0.5788831114768982, |
| "logits/rejected": -0.5802679061889648, |
| "logps/chosen": -12.416455268859863, |
| "logps/rejected": -15.200132369995117, |
| "loss": 1.3435254096984863, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5808812379837036, |
| "rewards/accuracies": 0.3125, |
| "rewards/chosen": 0.11838358640670776, |
| "rewards/margins": -0.03334012255072594, |
| "rewards/rejected": 0.1517237275838852, |
| "step": 21, |
| "train_speed(iter/s)": 0.005654 |
| }, |
| { |
| "epoch": 0.05262764446437916, |
| "grad_norm": 3.07053542137146, |
| "learning_rate": 6.984126984126984e-05, |
| "logits/chosen": -0.5787701606750488, |
| "logits/rejected": -0.582938551902771, |
| "logps/chosen": -6.899454116821289, |
| "logps/rejected": -21.719806671142578, |
| "loss": 1.0226894617080688, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.45206937193870544, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.30141958594322205, |
| "rewards/margins": 0.32421889901161194, |
| "rewards/rejected": -0.022799327969551086, |
| "step": 22, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 0.05501981012185094, |
| "grad_norm": 3.616273880004883, |
| "learning_rate": 7.301587301587302e-05, |
| "logits/chosen": -0.5814599990844727, |
| "logits/rejected": -0.5829823017120361, |
| "logps/chosen": -10.043213844299316, |
| "logps/rejected": -16.569318771362305, |
| "loss": 1.5021785497665405, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.8470967411994934, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.15741629898548126, |
| "rewards/margins": 0.13470463454723358, |
| "rewards/rejected": 0.022711653262376785, |
| "step": 23, |
| "train_speed(iter/s)": 0.005652 |
| }, |
| { |
| "epoch": 0.05741197577932272, |
| "grad_norm": 2.9444239139556885, |
| "learning_rate": 7.619047619047618e-05, |
| "logits/chosen": -0.5737470388412476, |
| "logits/rejected": -0.5753998756408691, |
| "logps/chosen": -9.754992485046387, |
| "logps/rejected": -23.393245697021484, |
| "loss": 1.1848090887069702, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5859429240226746, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.26539331674575806, |
| "rewards/margins": 0.25131887197494507, |
| "rewards/rejected": 0.01407446339726448, |
| "step": 24, |
| "train_speed(iter/s)": 0.005651 |
| }, |
| { |
| "epoch": 0.059804141436794495, |
| "grad_norm": 2.368903875350952, |
| "learning_rate": 7.936507936507937e-05, |
| "logits/chosen": -0.5797973871231079, |
| "logits/rejected": -0.5811963081359863, |
| "logps/chosen": -7.4247236251831055, |
| "logps/rejected": -17.2625789642334, |
| "loss": 1.1369986534118652, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5304021835327148, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.18115447461605072, |
| "rewards/margins": 0.21577578783035278, |
| "rewards/rejected": -0.034621305763721466, |
| "step": 25, |
| "train_speed(iter/s)": 0.005651 |
| }, |
| { |
| "epoch": 0.062196307094266275, |
| "grad_norm": 2.0124542713165283, |
| "learning_rate": 8.253968253968255e-05, |
| "logits/chosen": -0.5834470987319946, |
| "logits/rejected": -0.5854894518852234, |
| "logps/chosen": -10.883503913879395, |
| "logps/rejected": -17.59324836730957, |
| "loss": 1.050497055053711, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.4383053183555603, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.22693531215190887, |
| "rewards/margins": 0.23587419092655182, |
| "rewards/rejected": -0.008938884362578392, |
| "step": 26, |
| "train_speed(iter/s)": 0.005654 |
| }, |
| { |
| "epoch": 0.06458847275173805, |
| "grad_norm": 2.3302695751190186, |
| "learning_rate": 8.571428571428571e-05, |
| "logits/chosen": -0.5692424774169922, |
| "logits/rejected": -0.5711921453475952, |
| "logps/chosen": -12.02816104888916, |
| "logps/rejected": -23.778608322143555, |
| "loss": 1.1506422758102417, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.5187042951583862, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.11056558787822723, |
| "rewards/margins": 0.18688789010047913, |
| "rewards/rejected": -0.0763222873210907, |
| "step": 27, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 0.06698063840920984, |
| "grad_norm": 2.4422554969787598, |
| "learning_rate": 8.888888888888889e-05, |
| "logits/chosen": -0.5987146496772766, |
| "logits/rejected": -0.5993437767028809, |
| "logps/chosen": -9.00568675994873, |
| "logps/rejected": -17.24131965637207, |
| "loss": 1.0437039136886597, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.4295322299003601, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.19704687595367432, |
| "rewards/margins": 0.21514004468917847, |
| "rewards/rejected": -0.018093157559633255, |
| "step": 28, |
| "train_speed(iter/s)": 0.00565 |
| }, |
| { |
| "epoch": 0.06937280406668161, |
| "grad_norm": 3.3448095321655273, |
| "learning_rate": 9.206349206349206e-05, |
| "logits/chosen": -0.5584565997123718, |
| "logits/rejected": -0.5605788826942444, |
| "logps/chosen": -7.835123538970947, |
| "logps/rejected": -19.747636795043945, |
| "loss": 1.172892451286316, |
| "memory(GiB)": 46.29, |
| "nll_loss": 0.55903559923172, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.15631772577762604, |
| "rewards/margins": 0.2727695107460022, |
| "rewards/rejected": -0.11645177006721497, |
| "step": 29, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 0.0717649697241534, |
| "grad_norm": 2.956876516342163, |
| "learning_rate": 9.523809523809524e-05, |
| "logits/chosen": -0.5594294667243958, |
| "logits/rejected": -0.5630677342414856, |
| "logps/chosen": -8.298831939697266, |
| "logps/rejected": -26.14436149597168, |
| "loss": 1.0785880088806152, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5043998956680298, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.15360432863235474, |
| "rewards/margins": 0.37323492765426636, |
| "rewards/rejected": -0.21963059902191162, |
| "step": 30, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 0.07415713538162517, |
| "grad_norm": 2.969925880432129, |
| "learning_rate": 9.841269841269841e-05, |
| "logits/chosen": -0.563766360282898, |
| "logits/rejected": -0.564283549785614, |
| "logps/chosen": -9.237369537353516, |
| "logps/rejected": -23.180709838867188, |
| "loss": 1.0209345817565918, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.42502841353416443, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.18122270703315735, |
| "rewards/margins": 0.2958879768848419, |
| "rewards/rejected": -0.11466525495052338, |
| "step": 31, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 0.07654930103909696, |
| "grad_norm": 3.5162458419799805, |
| "learning_rate": 0.00010158730158730159, |
| "logits/chosen": -0.5594556927680969, |
| "logits/rejected": -0.5654793977737427, |
| "logps/chosen": -5.732161045074463, |
| "logps/rejected": -32.01944351196289, |
| "loss": 0.9676632881164551, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4746396541595459, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2609599530696869, |
| "rewards/margins": 0.6239939332008362, |
| "rewards/rejected": -0.3630340099334717, |
| "step": 32, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 0.07894146669656874, |
| "grad_norm": 3.3218283653259277, |
| "learning_rate": 0.00010476190476190477, |
| "logits/chosen": -0.5853260159492493, |
| "logits/rejected": -0.5895906686782837, |
| "logps/chosen": -10.368348121643066, |
| "logps/rejected": -20.515506744384766, |
| "loss": 1.1254382133483887, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5297057032585144, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.15998664498329163, |
| "rewards/margins": 0.346771240234375, |
| "rewards/rejected": -0.18678458034992218, |
| "step": 33, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 0.08133363235404052, |
| "grad_norm": 4.339278697967529, |
| "learning_rate": 0.00010793650793650794, |
| "logits/chosen": -0.6093581318855286, |
| "logits/rejected": -0.6088329553604126, |
| "logps/chosen": -13.336594581604004, |
| "logps/rejected": -24.60618782043457, |
| "loss": 1.1337980031967163, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5539220571517944, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.1875077337026596, |
| "rewards/margins": 0.4340534806251526, |
| "rewards/rejected": -0.24654576182365417, |
| "step": 34, |
| "train_speed(iter/s)": 0.005642 |
| }, |
| { |
| "epoch": 0.0837257980115123, |
| "grad_norm": 3.8656299114227295, |
| "learning_rate": 0.00011111111111111112, |
| "logits/chosen": -0.6122190952301025, |
| "logits/rejected": -0.616493284702301, |
| "logps/chosen": -13.914678573608398, |
| "logps/rejected": -32.916744232177734, |
| "loss": 1.1174585819244385, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5647247433662415, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.015515260398387909, |
| "rewards/margins": 0.6349303722381592, |
| "rewards/rejected": -0.6194151639938354, |
| "step": 35, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.08611796366898408, |
| "grad_norm": 3.665461540222168, |
| "learning_rate": 0.00011428571428571428, |
| "logits/chosen": -0.6845042705535889, |
| "logits/rejected": -0.6848981380462646, |
| "logps/chosen": -12.846878051757812, |
| "logps/rejected": -25.240447998046875, |
| "loss": 1.1161198616027832, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5125420689582825, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.049385204911231995, |
| "rewards/margins": 0.33047205209732056, |
| "rewards/rejected": -0.28108683228492737, |
| "step": 36, |
| "train_speed(iter/s)": 0.005643 |
| }, |
| { |
| "epoch": 0.08851012932645586, |
| "grad_norm": 3.6945273876190186, |
| "learning_rate": 0.00011746031746031746, |
| "logits/chosen": -0.6884023547172546, |
| "logits/rejected": -0.69326251745224, |
| "logps/chosen": -7.306849956512451, |
| "logps/rejected": -24.99851417541504, |
| "loss": 0.9112467765808105, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4103700518608093, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.24825294315814972, |
| "rewards/margins": 0.5438191294670105, |
| "rewards/rejected": -0.295566201210022, |
| "step": 37, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 0.09090229498392764, |
| "grad_norm": 2.759567975997925, |
| "learning_rate": 0.00012063492063492063, |
| "logits/chosen": -0.6944407224655151, |
| "logits/rejected": -0.6986436247825623, |
| "logps/chosen": -8.414443969726562, |
| "logps/rejected": -29.627033233642578, |
| "loss": 0.9807862043380737, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.464620977640152, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.205428808927536, |
| "rewards/margins": 0.6206997036933899, |
| "rewards/rejected": -0.41527092456817627, |
| "step": 38, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 0.09329446064139942, |
| "grad_norm": 3.707176685333252, |
| "learning_rate": 0.0001238095238095238, |
| "logits/chosen": -0.6549853086471558, |
| "logits/rejected": -0.6607218384742737, |
| "logps/chosen": -13.402036666870117, |
| "logps/rejected": -29.2238826751709, |
| "loss": 1.1863672733306885, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5955373048782349, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.0599823072552681, |
| "rewards/margins": 0.4598371684551239, |
| "rewards/rejected": -0.3998548984527588, |
| "step": 39, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 0.0956866262988712, |
| "grad_norm": 5.015953540802002, |
| "learning_rate": 0.00012698412698412698, |
| "logits/chosen": -0.6966100931167603, |
| "logits/rejected": -0.7002891302108765, |
| "logps/chosen": -7.736152172088623, |
| "logps/rejected": -27.045543670654297, |
| "loss": 0.977799654006958, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5642723441123962, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.3473990261554718, |
| "rewards/margins": 1.0138760805130005, |
| "rewards/rejected": -0.6664770245552063, |
| "step": 40, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 0.09807879195634298, |
| "grad_norm": 4.407506465911865, |
| "learning_rate": 0.00013015873015873017, |
| "logits/chosen": -0.692713737487793, |
| "logits/rejected": -0.6956310272216797, |
| "logps/chosen": -9.848102569580078, |
| "logps/rejected": -24.158405303955078, |
| "loss": 1.0173840522766113, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5190764665603638, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1344590187072754, |
| "rewards/margins": 0.6329586505889893, |
| "rewards/rejected": -0.49849969148635864, |
| "step": 41, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 0.10047095761381476, |
| "grad_norm": 3.482964515686035, |
| "learning_rate": 0.00013333333333333334, |
| "logits/chosen": -0.6898783445358276, |
| "logits/rejected": -0.694994330406189, |
| "logps/chosen": -10.68550968170166, |
| "logps/rejected": -28.72760772705078, |
| "loss": 1.0436055660247803, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5537880659103394, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.1501934826374054, |
| "rewards/margins": 0.8055344820022583, |
| "rewards/rejected": -0.6553410291671753, |
| "step": 42, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 0.10286312327128654, |
| "grad_norm": 3.4329941272735596, |
| "learning_rate": 0.0001365079365079365, |
| "logits/chosen": -0.7011124491691589, |
| "logits/rejected": -0.7064176797866821, |
| "logps/chosen": -15.171253204345703, |
| "logps/rejected": -29.2159481048584, |
| "loss": 1.0602749586105347, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5264595746994019, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.06841089576482773, |
| "rewards/margins": 0.6964786052703857, |
| "rewards/rejected": -0.6280677318572998, |
| "step": 43, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 0.10525528892875832, |
| "grad_norm": 2.459496259689331, |
| "learning_rate": 0.00013968253968253967, |
| "logits/chosen": -0.7666099667549133, |
| "logits/rejected": -0.771510899066925, |
| "logps/chosen": -5.71858549118042, |
| "logps/rejected": -27.757709503173828, |
| "loss": 0.8418532609939575, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.43147915601730347, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2912772297859192, |
| "rewards/margins": 0.8343029022216797, |
| "rewards/rejected": -0.5430256724357605, |
| "step": 44, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 0.1076474545862301, |
| "grad_norm": 2.439218759536743, |
| "learning_rate": 0.00014285714285714287, |
| "logits/chosen": -0.7568098902702332, |
| "logits/rejected": -0.7595718502998352, |
| "logps/chosen": -10.82071304321289, |
| "logps/rejected": -26.015369415283203, |
| "loss": 0.9067925214767456, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4222971796989441, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.23031610250473022, |
| "rewards/margins": 0.6913963556289673, |
| "rewards/rejected": -0.46108025312423706, |
| "step": 45, |
| "train_speed(iter/s)": 0.005643 |
| }, |
| { |
| "epoch": 0.11003962024370188, |
| "grad_norm": 2.0914173126220703, |
| "learning_rate": 0.00014603174603174603, |
| "logits/chosen": -0.8171934485435486, |
| "logits/rejected": -0.8208147287368774, |
| "logps/chosen": -7.6959333419799805, |
| "logps/rejected": -26.958995819091797, |
| "loss": 0.8884089589118958, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4511643052101135, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.29817235469818115, |
| "rewards/margins": 0.9990655183792114, |
| "rewards/rejected": -0.700893223285675, |
| "step": 46, |
| "train_speed(iter/s)": 0.005641 |
| }, |
| { |
| "epoch": 0.11243178590117366, |
| "grad_norm": 2.7921295166015625, |
| "learning_rate": 0.00014920634920634923, |
| "logits/chosen": -0.8353937864303589, |
| "logits/rejected": -0.8423863649368286, |
| "logps/chosen": -11.051738739013672, |
| "logps/rejected": -29.439268112182617, |
| "loss": 0.9526035189628601, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5338684916496277, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.15457165241241455, |
| "rewards/margins": 0.8811558485031128, |
| "rewards/rejected": -0.7265841960906982, |
| "step": 47, |
| "train_speed(iter/s)": 0.00564 |
| }, |
| { |
| "epoch": 0.11482395155864544, |
| "grad_norm": 3.267803907394409, |
| "learning_rate": 0.00015238095238095237, |
| "logits/chosen": -0.8806051015853882, |
| "logits/rejected": -0.8819608688354492, |
| "logps/chosen": -15.4578275680542, |
| "logps/rejected": -28.783483505249023, |
| "loss": 1.037742018699646, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5505475997924805, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.205413818359375, |
| "rewards/margins": 0.9907757043838501, |
| "rewards/rejected": -0.7853618264198303, |
| "step": 48, |
| "train_speed(iter/s)": 0.005641 |
| }, |
| { |
| "epoch": 0.11721611721611722, |
| "grad_norm": 5.174523830413818, |
| "learning_rate": 0.00015555555555555556, |
| "logits/chosen": -0.9132509827613831, |
| "logits/rejected": -0.9135016202926636, |
| "logps/chosen": -16.307157516479492, |
| "logps/rejected": -32.103309631347656, |
| "loss": 1.1124699115753174, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6212166547775269, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.16737183928489685, |
| "rewards/margins": 0.9279904365539551, |
| "rewards/rejected": -0.7606185078620911, |
| "step": 49, |
| "train_speed(iter/s)": 0.005639 |
| }, |
| { |
| "epoch": 0.11960828287358899, |
| "grad_norm": 3.378316879272461, |
| "learning_rate": 0.00015873015873015873, |
| "logits/chosen": -0.9473586678504944, |
| "logits/rejected": -0.958495020866394, |
| "logps/chosen": -7.192904472351074, |
| "logps/rejected": -38.12461853027344, |
| "loss": 0.8130012154579163, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4022335410118103, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.14782631397247314, |
| "rewards/margins": 1.2623234987258911, |
| "rewards/rejected": -1.114497184753418, |
| "step": 50, |
| "train_speed(iter/s)": 0.005639 |
| }, |
| { |
| "epoch": 0.12200044853106078, |
| "grad_norm": 3.7333006858825684, |
| "learning_rate": 0.00016190476190476192, |
| "logits/chosen": -0.9992814064025879, |
| "logits/rejected": -1.0156452655792236, |
| "logps/chosen": -10.031129837036133, |
| "logps/rejected": -41.15076446533203, |
| "loss": 0.8442965745925903, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.538231611251831, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.17759491503238678, |
| "rewards/margins": 1.8191030025482178, |
| "rewards/rejected": -1.6415079832077026, |
| "step": 51, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.12439261418853255, |
| "grad_norm": 5.16239070892334, |
| "learning_rate": 0.0001650793650793651, |
| "logits/chosen": -1.0308305025100708, |
| "logits/rejected": -1.0365955829620361, |
| "logps/chosen": -12.618029594421387, |
| "logps/rejected": -30.713321685791016, |
| "loss": 1.0460460186004639, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6516051292419434, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.07658500969409943, |
| "rewards/margins": 1.1642749309539795, |
| "rewards/rejected": -1.0876898765563965, |
| "step": 52, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.12678477984600434, |
| "grad_norm": 8.370670318603516, |
| "learning_rate": 0.00016825396825396826, |
| "logits/chosen": -0.9728441834449768, |
| "logits/rejected": -0.9805364012718201, |
| "logps/chosen": -15.092721939086914, |
| "logps/rejected": -43.16249465942383, |
| "loss": 1.3046923875808716, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.8397946953773499, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.006958425045013428, |
| "rewards/margins": 1.5226078033447266, |
| "rewards/rejected": -1.5295664072036743, |
| "step": 53, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.1291769455034761, |
| "grad_norm": 4.0541462898254395, |
| "learning_rate": 0.00017142857142857143, |
| "logits/chosen": -1.0004466772079468, |
| "logits/rejected": -1.0036160945892334, |
| "logps/chosen": -10.588321685791016, |
| "logps/rejected": -31.380657196044922, |
| "loss": 0.784751296043396, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.43249669671058655, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.19054383039474487, |
| "rewards/margins": 1.5502821207046509, |
| "rewards/rejected": -1.3597382307052612, |
| "step": 54, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.13156911116094788, |
| "grad_norm": 5.073096752166748, |
| "learning_rate": 0.00017460317460317462, |
| "logits/chosen": -1.0498322248458862, |
| "logits/rejected": -1.0584659576416016, |
| "logps/chosen": -15.89664363861084, |
| "logps/rejected": -31.34156608581543, |
| "loss": 1.277337908744812, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.8222866654396057, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.005467001348733902, |
| "rewards/margins": 1.1543742418289185, |
| "rewards/rejected": -1.1598412990570068, |
| "step": 55, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.13396127681841968, |
| "grad_norm": 9.963640213012695, |
| "learning_rate": 0.00017777777777777779, |
| "logits/chosen": -0.9661000370979309, |
| "logits/rejected": -0.9783132672309875, |
| "logps/chosen": -10.20169448852539, |
| "logps/rejected": -36.709903717041016, |
| "loss": 1.0616700649261475, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6122196912765503, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.12718632817268372, |
| "rewards/margins": 1.107292890548706, |
| "rewards/rejected": -0.9801065921783447, |
| "step": 56, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.13635344247589146, |
| "grad_norm": 4.3275227546691895, |
| "learning_rate": 0.00018095238095238095, |
| "logits/chosen": -1.0224848985671997, |
| "logits/rejected": -1.0346163511276245, |
| "logps/chosen": -12.20643424987793, |
| "logps/rejected": -46.608978271484375, |
| "loss": 0.8705050349235535, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.501814067363739, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.06085830181837082, |
| "rewards/margins": 1.5555297136306763, |
| "rewards/rejected": -1.49467134475708, |
| "step": 57, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.13874560813336323, |
| "grad_norm": 7.873358726501465, |
| "learning_rate": 0.00018412698412698412, |
| "logits/chosen": -0.9996501207351685, |
| "logits/rejected": -1.008366346359253, |
| "logps/chosen": -8.571969985961914, |
| "logps/rejected": -36.11326599121094, |
| "loss": 0.8140489459037781, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4696846008300781, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.10605719685554504, |
| "rewards/margins": 1.5816770792007446, |
| "rewards/rejected": -1.475619912147522, |
| "step": 58, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.141137773790835, |
| "grad_norm": 4.191744804382324, |
| "learning_rate": 0.00018730158730158731, |
| "logits/chosen": -0.9680933952331543, |
| "logits/rejected": -0.9731469750404358, |
| "logps/chosen": -11.40235710144043, |
| "logps/rejected": -45.129981994628906, |
| "loss": 0.8071048259735107, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.45748206973075867, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.023274360224604607, |
| "rewards/margins": 2.327709913253784, |
| "rewards/rejected": -2.3509843349456787, |
| "step": 59, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.1435299394483068, |
| "grad_norm": 3.393984317779541, |
| "learning_rate": 0.00019047619047619048, |
| "logits/chosen": -0.9939959645271301, |
| "logits/rejected": -0.9985793232917786, |
| "logps/chosen": -10.470319747924805, |
| "logps/rejected": -34.314205169677734, |
| "loss": 0.8613054156303406, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5560427904129028, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.24216137826442719, |
| "rewards/margins": 1.7053587436676025, |
| "rewards/rejected": -1.4631972312927246, |
| "step": 60, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.14592210510577858, |
| "grad_norm": 2.9272122383117676, |
| "learning_rate": 0.00019365079365079365, |
| "logits/chosen": -0.9692378044128418, |
| "logits/rejected": -0.9806127548217773, |
| "logps/chosen": -9.072213172912598, |
| "logps/rejected": -40.46186828613281, |
| "loss": 0.7413874864578247, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.42558133602142334, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.12333947420120239, |
| "rewards/margins": 1.8263578414916992, |
| "rewards/rejected": -1.703018307685852, |
| "step": 61, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.14831427076325035, |
| "grad_norm": 6.832886219024658, |
| "learning_rate": 0.00019682539682539682, |
| "logits/chosen": -0.9575268030166626, |
| "logits/rejected": -0.9714451432228088, |
| "logps/chosen": -12.573237419128418, |
| "logps/rejected": -46.78509521484375, |
| "loss": 1.1069118976593018, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.7730302214622498, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.055254869163036346, |
| "rewards/margins": 1.9246735572814941, |
| "rewards/rejected": -1.9799282550811768, |
| "step": 62, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.15070643642072212, |
| "grad_norm": 3.042998790740967, |
| "learning_rate": 0.0002, |
| "logits/chosen": -0.982741117477417, |
| "logits/rejected": -0.9914990663528442, |
| "logps/chosen": -7.9191131591796875, |
| "logps/rejected": -37.919273376464844, |
| "loss": 0.7625957131385803, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.39874428510665894, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.21871888637542725, |
| "rewards/margins": 1.7384941577911377, |
| "rewards/rejected": -1.5197750329971313, |
| "step": 63, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.15309860207819392, |
| "grad_norm": 4.333725452423096, |
| "learning_rate": 0.00019999965210677203, |
| "logits/chosen": -0.923725962638855, |
| "logits/rejected": -0.9260438084602356, |
| "logps/chosen": -10.865621566772461, |
| "logps/rejected": -28.435291290283203, |
| "loss": 1.0750609636306763, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5927442908287048, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.0741436630487442, |
| "rewards/margins": 0.9901498556137085, |
| "rewards/rejected": -0.9160062074661255, |
| "step": 64, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.1554907677356657, |
| "grad_norm": 3.861036539077759, |
| "learning_rate": 0.00019999860842950868, |
| "logits/chosen": -0.9479286670684814, |
| "logits/rejected": -0.9511004686355591, |
| "logps/chosen": -14.088654518127441, |
| "logps/rejected": -34.085487365722656, |
| "loss": 0.8642847537994385, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4594627618789673, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.02415481209754944, |
| "rewards/margins": 1.243844747543335, |
| "rewards/rejected": -1.2679994106292725, |
| "step": 65, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.15788293339313747, |
| "grad_norm": 2.964956283569336, |
| "learning_rate": 0.00019999686897547169, |
| "logits/chosen": -0.8965855240821838, |
| "logits/rejected": -0.9040545225143433, |
| "logps/chosen": -11.898802757263184, |
| "logps/rejected": -30.001935958862305, |
| "loss": 0.9386196136474609, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5237320065498352, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.20598752796649933, |
| "rewards/margins": 1.1080772876739502, |
| "rewards/rejected": -0.9020897150039673, |
| "step": 66, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.16027509905060924, |
| "grad_norm": 2.645886182785034, |
| "learning_rate": 0.000199994433756764, |
| "logits/chosen": -0.9297692775726318, |
| "logits/rejected": -0.9331662058830261, |
| "logps/chosen": -9.838333129882812, |
| "logps/rejected": -31.623937606811523, |
| "loss": 0.9248643517494202, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5388163328170776, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.30869799852371216, |
| "rewards/margins": 1.5470086336135864, |
| "rewards/rejected": -1.2383108139038086, |
| "step": 67, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.16266726470808104, |
| "grad_norm": 4.206631660461426, |
| "learning_rate": 0.00019999130279032952, |
| "logits/chosen": -0.8769949674606323, |
| "logits/rejected": -0.8823909759521484, |
| "logps/chosen": -18.815860748291016, |
| "logps/rejected": -40.8411865234375, |
| "loss": 1.2017946243286133, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.7082366347312927, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.15441593527793884, |
| "rewards/margins": 1.270005702972412, |
| "rewards/rejected": -1.4244215488433838, |
| "step": 68, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.16505943036555282, |
| "grad_norm": 2.9063844680786133, |
| "learning_rate": 0.00019998747609795308, |
| "logits/chosen": -0.9327405691146851, |
| "logits/rejected": -0.9352924823760986, |
| "logps/chosen": -8.574593544006348, |
| "logps/rejected": -26.808048248291016, |
| "loss": 0.9598045349121094, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.48059219121932983, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.14917421340942383, |
| "rewards/margins": 1.1800099611282349, |
| "rewards/rejected": -1.0308358669281006, |
| "step": 69, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.1674515960230246, |
| "grad_norm": 3.173595428466797, |
| "learning_rate": 0.00019998295370626026, |
| "logits/chosen": -0.9241899847984314, |
| "logits/rejected": -0.9277300834655762, |
| "logps/chosen": -8.667724609375, |
| "logps/rejected": -33.7216911315918, |
| "loss": 0.7956899404525757, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.46254533529281616, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.3548792004585266, |
| "rewards/margins": 1.690018653869629, |
| "rewards/rejected": -1.3351393938064575, |
| "step": 70, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.16984376168049636, |
| "grad_norm": 2.7610597610473633, |
| "learning_rate": 0.00019997773564671732, |
| "logits/chosen": -0.9446437358856201, |
| "logits/rejected": -0.948798656463623, |
| "logps/chosen": -7.366062164306641, |
| "logps/rejected": -33.738685607910156, |
| "loss": 0.7108196020126343, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.34850752353668213, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.19515307247638702, |
| "rewards/margins": 1.521824836730957, |
| "rewards/rejected": -1.3266719579696655, |
| "step": 71, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.17223592733796816, |
| "grad_norm": 2.687474250793457, |
| "learning_rate": 0.00019997182195563072, |
| "logits/chosen": -0.9068199396133423, |
| "logits/rejected": -0.9117379188537598, |
| "logps/chosen": -11.47612476348877, |
| "logps/rejected": -36.21818161010742, |
| "loss": 0.7809538245201111, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4381601810455322, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.23770935833454132, |
| "rewards/margins": 1.4757200479507446, |
| "rewards/rejected": -1.2380106449127197, |
| "step": 72, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.17462809299543994, |
| "grad_norm": 4.630646228790283, |
| "learning_rate": 0.00019996521267414717, |
| "logits/chosen": -0.9765245914459229, |
| "logits/rejected": -0.9796733856201172, |
| "logps/chosen": -18.482873916625977, |
| "logps/rejected": -29.491580963134766, |
| "loss": 1.1687482595443726, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.703139066696167, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.22344376146793365, |
| "rewards/margins": 1.3359349966049194, |
| "rewards/rejected": -1.1124913692474365, |
| "step": 73, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.1770202586529117, |
| "grad_norm": 2.8936681747436523, |
| "learning_rate": 0.00019995790784825318, |
| "logits/chosen": -0.9390259385108948, |
| "logits/rejected": -0.9489994049072266, |
| "logps/chosen": -7.563755035400391, |
| "logps/rejected": -45.333255767822266, |
| "loss": 0.6042292714118958, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.29600095748901367, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2876776158809662, |
| "rewards/margins": 2.2362325191497803, |
| "rewards/rejected": -1.9485547542572021, |
| "step": 74, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.17941242431038348, |
| "grad_norm": 7.294376850128174, |
| "learning_rate": 0.00019994990752877472, |
| "logits/chosen": -0.9351106882095337, |
| "logits/rejected": -0.9450594186782837, |
| "logps/chosen": -8.70926570892334, |
| "logps/rejected": -40.92999267578125, |
| "loss": 0.8546091914176941, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4996742010116577, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.15468531847000122, |
| "rewards/margins": 1.8762543201446533, |
| "rewards/rejected": -1.7215690612792969, |
| "step": 75, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.18180458996785528, |
| "grad_norm": 3.972262382507324, |
| "learning_rate": 0.0001999412117713769, |
| "logits/chosen": -1.0458097457885742, |
| "logits/rejected": -1.0525654554367065, |
| "logps/chosen": -14.451887130737305, |
| "logps/rejected": -50.781612396240234, |
| "loss": 1.1873810291290283, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.7520004510879517, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.15246914327144623, |
| "rewards/margins": 1.8416060209274292, |
| "rewards/rejected": -1.9940751791000366, |
| "step": 76, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.18419675562532706, |
| "grad_norm": 4.1366729736328125, |
| "learning_rate": 0.00019993182063656365, |
| "logits/chosen": -0.9752227663993835, |
| "logits/rejected": -0.9808940887451172, |
| "logps/chosen": -13.090173721313477, |
| "logps/rejected": -46.20813751220703, |
| "loss": 0.9597315788269043, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.564085841178894, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.06369027495384216, |
| "rewards/margins": 2.183162212371826, |
| "rewards/rejected": -2.119471788406372, |
| "step": 77, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.18658892128279883, |
| "grad_norm": 7.812252044677734, |
| "learning_rate": 0.00019992173418967722, |
| "logits/chosen": -1.0128633975982666, |
| "logits/rejected": -1.0182214975357056, |
| "logps/chosen": -8.333160400390625, |
| "logps/rejected": -49.87608337402344, |
| "loss": 0.7557947635650635, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5276219248771667, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.05898723006248474, |
| "rewards/margins": 3.0209884643554688, |
| "rewards/rejected": -2.962001323699951, |
| "step": 78, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.1889810869402706, |
| "grad_norm": 5.186205863952637, |
| "learning_rate": 0.0001999109525008977, |
| "logits/chosen": -1.006530523300171, |
| "logits/rejected": -1.0176763534545898, |
| "logps/chosen": -16.32330894470215, |
| "logps/rejected": -42.1322135925293, |
| "loss": 1.0281399488449097, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6972037553787231, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.12587852776050568, |
| "rewards/margins": 2.273554801940918, |
| "rewards/rejected": -2.1476762294769287, |
| "step": 79, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.1913732525977424, |
| "grad_norm": 3.2977113723754883, |
| "learning_rate": 0.00019989947564524267, |
| "logits/chosen": -0.9115131497383118, |
| "logits/rejected": -0.9150525331497192, |
| "logps/chosen": -14.462236404418945, |
| "logps/rejected": -37.0766487121582, |
| "loss": 0.9416806697845459, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.618413507938385, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1239091008901596, |
| "rewards/margins": 1.5363889932632446, |
| "rewards/rejected": -1.4124797582626343, |
| "step": 80, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.19376541825521418, |
| "grad_norm": 2.5254037380218506, |
| "learning_rate": 0.00019988730370256653, |
| "logits/chosen": -0.923732340335846, |
| "logits/rejected": -0.9374791383743286, |
| "logps/chosen": -7.155979156494141, |
| "logps/rejected": -48.05502700805664, |
| "loss": 0.7125514149665833, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.3992992639541626, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.3067445456981659, |
| "rewards/margins": 2.08925461769104, |
| "rewards/rejected": -1.7825100421905518, |
| "step": 81, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.19615758391268595, |
| "grad_norm": 9.468795776367188, |
| "learning_rate": 0.00019987443675756, |
| "logits/chosen": -0.9325875043869019, |
| "logits/rejected": -0.9395610094070435, |
| "logps/chosen": -10.340322494506836, |
| "logps/rejected": -39.463584899902344, |
| "loss": 0.9755638241767883, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6087434887886047, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.17063137888908386, |
| "rewards/margins": 1.9463139772415161, |
| "rewards/rejected": -1.775682806968689, |
| "step": 82, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.19854974957015772, |
| "grad_norm": 3.6070539951324463, |
| "learning_rate": 0.00019986087489974952, |
| "logits/chosen": -0.953080952167511, |
| "logits/rejected": -0.9536972045898438, |
| "logps/chosen": -12.9752197265625, |
| "logps/rejected": -32.04649353027344, |
| "loss": 1.0608779191970825, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6077963709831238, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.20461547374725342, |
| "rewards/margins": 1.1314504146575928, |
| "rewards/rejected": -0.9268349409103394, |
| "step": 83, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.20094191522762953, |
| "grad_norm": 3.0724594593048096, |
| "learning_rate": 0.00019984661822349667, |
| "logits/chosen": -0.9149019718170166, |
| "logits/rejected": -0.9223135113716125, |
| "logps/chosen": -11.019997596740723, |
| "logps/rejected": -34.67535400390625, |
| "loss": 0.959977924823761, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5666155815124512, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.07833094894886017, |
| "rewards/margins": 1.359312891960144, |
| "rewards/rejected": -1.2809820175170898, |
| "step": 84, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.2033340808851013, |
| "grad_norm": 4.832602024078369, |
| "learning_rate": 0.0001998316668279975, |
| "logits/chosen": -0.9071481823921204, |
| "logits/rejected": -0.9128391742706299, |
| "logps/chosen": -10.688645362854004, |
| "logps/rejected": -35.94822311401367, |
| "loss": 0.8470439314842224, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4921894371509552, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.15914583206176758, |
| "rewards/margins": 1.6146492958068848, |
| "rewards/rejected": -1.4555033445358276, |
| "step": 85, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.20572624654257307, |
| "grad_norm": 2.5131633281707764, |
| "learning_rate": 0.00019981602081728178, |
| "logits/chosen": -0.9089372754096985, |
| "logits/rejected": -0.9133983850479126, |
| "logps/chosen": -9.597051620483398, |
| "logps/rejected": -31.3452205657959, |
| "loss": 0.8543159365653992, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.49588966369628906, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.37841910123825073, |
| "rewards/margins": 1.7511966228485107, |
| "rewards/rejected": -1.3727774620056152, |
| "step": 86, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.20811841220004484, |
| "grad_norm": 2.388970136642456, |
| "learning_rate": 0.0001997996803002123, |
| "logits/chosen": -0.9073879718780518, |
| "logits/rejected": -0.9152944684028625, |
| "logps/chosen": -9.81510066986084, |
| "logps/rejected": -44.42865753173828, |
| "loss": 0.7038865685462952, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4832083582878113, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.40937453508377075, |
| "rewards/margins": 2.5496997833251953, |
| "rewards/rejected": -2.1403250694274902, |
| "step": 87, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.21051057785751665, |
| "grad_norm": 4.166887283325195, |
| "learning_rate": 0.0001997826453904842, |
| "logits/chosen": -0.9330130815505981, |
| "logits/rejected": -0.9432638883590698, |
| "logps/chosen": -8.299344062805176, |
| "logps/rejected": -45.63969421386719, |
| "loss": 0.8084508180618286, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5791301727294922, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.28511884808540344, |
| "rewards/margins": 2.360887050628662, |
| "rewards/rejected": -2.07576847076416, |
| "step": 88, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.21290274351498842, |
| "grad_norm": 2.945251703262329, |
| "learning_rate": 0.00019976491620662406, |
| "logits/chosen": -0.9447211623191833, |
| "logits/rejected": -0.9499524235725403, |
| "logps/chosen": -11.17585277557373, |
| "logps/rejected": -39.13389205932617, |
| "loss": 0.8930118680000305, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5336206555366516, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2357935607433319, |
| "rewards/margins": 1.8502506017684937, |
| "rewards/rejected": -1.6144568920135498, |
| "step": 89, |
| "train_speed(iter/s)": 0.005651 |
| }, |
| { |
| "epoch": 0.2152949091724602, |
| "grad_norm": 2.779130697250366, |
| "learning_rate": 0.00019974649287198917, |
| "logits/chosen": -0.9709842205047607, |
| "logits/rejected": -0.9703356623649597, |
| "logps/chosen": -14.912643432617188, |
| "logps/rejected": -38.99196243286133, |
| "loss": 0.8809282779693604, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5660299062728882, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.05061885714530945, |
| "rewards/margins": 2.167189598083496, |
| "rewards/rejected": -2.116570472717285, |
| "step": 90, |
| "train_speed(iter/s)": 0.00565 |
| }, |
| { |
| "epoch": 0.21768707482993196, |
| "grad_norm": 18.56264305114746, |
| "learning_rate": 0.00019972737551476657, |
| "logits/chosen": -0.9611991047859192, |
| "logits/rejected": -0.9779796600341797, |
| "logps/chosen": -12.885236740112305, |
| "logps/rejected": -61.122474670410156, |
| "loss": 1.059419870376587, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.7570096254348755, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.2121628373861313, |
| "rewards/margins": 3.2702770233154297, |
| "rewards/rejected": -3.4824395179748535, |
| "step": 91, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 0.22007924048740377, |
| "grad_norm": 12.53586196899414, |
| "learning_rate": 0.00019970756426797222, |
| "logits/chosen": -0.9786768555641174, |
| "logits/rejected": -0.9855670928955078, |
| "logps/chosen": -11.347864151000977, |
| "logps/rejected": -53.41314697265625, |
| "loss": 1.0533485412597656, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.7857727408409119, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.0066121164709329605, |
| "rewards/margins": 3.5626165866851807, |
| "rewards/rejected": -3.5692288875579834, |
| "step": 92, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 0.22247140614487554, |
| "grad_norm": 5.693381309509277, |
| "learning_rate": 0.00019968705926945015, |
| "logits/chosen": -0.9693127870559692, |
| "logits/rejected": -0.9756276607513428, |
| "logps/chosen": -12.194255828857422, |
| "logps/rejected": -57.608524322509766, |
| "loss": 0.8848276138305664, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5927705764770508, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.047529980540275574, |
| "rewards/margins": 3.575960397720337, |
| "rewards/rejected": -3.623490333557129, |
| "step": 93, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 0.2248635718023473, |
| "grad_norm": 6.475986957550049, |
| "learning_rate": 0.00019966586066187133, |
| "logits/chosen": -0.9738945960998535, |
| "logits/rejected": -0.9793066382408142, |
| "logps/chosen": -13.28403377532959, |
| "logps/rejected": -44.68003845214844, |
| "loss": 1.1366519927978516, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6431254148483276, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.027368444949388504, |
| "rewards/margins": 2.1656975746154785, |
| "rewards/rejected": -2.193066120147705, |
| "step": 94, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 0.22725573745981908, |
| "grad_norm": 3.099266529083252, |
| "learning_rate": 0.00019964396859273277, |
| "logits/chosen": -0.911146879196167, |
| "logits/rejected": -0.918366551399231, |
| "logps/chosen": -11.675085067749023, |
| "logps/rejected": -43.39706802368164, |
| "loss": 1.0351701974868774, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6883416175842285, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.11412099003791809, |
| "rewards/margins": 2.480591297149658, |
| "rewards/rejected": -2.3664703369140625, |
| "step": 95, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 0.22964790311729089, |
| "grad_norm": 2.586851119995117, |
| "learning_rate": 0.00019962138321435657, |
| "logits/chosen": -0.8519182205200195, |
| "logits/rejected": -0.8646640181541443, |
| "logps/chosen": -9.719616889953613, |
| "logps/rejected": -47.84657669067383, |
| "loss": 0.7686633467674255, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5021700859069824, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.3157491087913513, |
| "rewards/margins": 2.579385280609131, |
| "rewards/rejected": -2.2636361122131348, |
| "step": 96, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 0.23204006877476266, |
| "grad_norm": 2.8380484580993652, |
| "learning_rate": 0.0001995981046838887, |
| "logits/chosen": -0.8343847990036011, |
| "logits/rejected": -0.8459703326225281, |
| "logps/chosen": -18.732688903808594, |
| "logps/rejected": -45.6234016418457, |
| "loss": 0.8623945116996765, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5681560635566711, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.1901763379573822, |
| "rewards/margins": 2.299340009689331, |
| "rewards/rejected": -2.109163522720337, |
| "step": 97, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 0.23443223443223443, |
| "grad_norm": 4.969798564910889, |
| "learning_rate": 0.00019957413316329807, |
| "logits/chosen": -0.747201681137085, |
| "logits/rejected": -0.7517687082290649, |
| "logps/chosen": -17.974029541015625, |
| "logps/rejected": -45.19165802001953, |
| "loss": 1.0327142477035522, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5473703145980835, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.12275683879852295, |
| "rewards/margins": 2.03201961517334, |
| "rewards/rejected": -2.1547763347625732, |
| "step": 98, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 0.2368244000897062, |
| "grad_norm": 4.590868949890137, |
| "learning_rate": 0.00019954946881937523, |
| "logits/chosen": -0.7821736335754395, |
| "logits/rejected": -0.7837796807289124, |
| "logps/chosen": -10.349934577941895, |
| "logps/rejected": -31.049577713012695, |
| "loss": 0.8570771217346191, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5170512199401855, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.12493617832660675, |
| "rewards/margins": 1.5637221336364746, |
| "rewards/rejected": -1.4387857913970947, |
| "step": 99, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 0.23921656574717798, |
| "grad_norm": 3.0355851650238037, |
| "learning_rate": 0.00019952411182373137, |
| "logits/chosen": -0.7416669726371765, |
| "logits/rejected": -0.7514837980270386, |
| "logps/chosen": -12.758575439453125, |
| "logps/rejected": -52.49340057373047, |
| "loss": 0.7629787921905518, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4688795804977417, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.32235589623451233, |
| "rewards/margins": 2.829197406768799, |
| "rewards/rejected": -2.5068414211273193, |
| "step": 100, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 0.24160873140464978, |
| "grad_norm": 4.2240214347839355, |
| "learning_rate": 0.000199498062352797, |
| "logits/chosen": -0.7358923554420471, |
| "logits/rejected": -0.7415207624435425, |
| "logps/chosen": -12.469839096069336, |
| "logps/rejected": -39.323734283447266, |
| "loss": 0.8776499629020691, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5749576091766357, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2892524003982544, |
| "rewards/margins": 1.7897101640701294, |
| "rewards/rejected": -1.5004578828811646, |
| "step": 101, |
| "train_speed(iter/s)": 0.00564 |
| }, |
| { |
| "epoch": 0.24400089706212155, |
| "grad_norm": 4.133398532867432, |
| "learning_rate": 0.00019947132058782083, |
| "logits/chosen": -0.7967073321342468, |
| "logits/rejected": -0.7993899583816528, |
| "logps/chosen": -9.669886589050293, |
| "logps/rejected": -25.658836364746094, |
| "loss": 0.9530386924743652, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5376937985420227, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.09255746752023697, |
| "rewards/margins": 1.1425445079803467, |
| "rewards/rejected": -1.0499870777130127, |
| "step": 102, |
| "train_speed(iter/s)": 0.00564 |
| }, |
| { |
| "epoch": 0.24639306271959333, |
| "grad_norm": 4.5498857498168945, |
| "learning_rate": 0.00019944388671486844, |
| "logits/chosen": -0.7290014028549194, |
| "logits/rejected": -0.7364125847816467, |
| "logps/chosen": -10.035048484802246, |
| "logps/rejected": -38.11687088012695, |
| "loss": 0.8911282420158386, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4538458585739136, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.18342246115207672, |
| "rewards/margins": 1.3529313802719116, |
| "rewards/rejected": -1.169508934020996, |
| "step": 103, |
| "train_speed(iter/s)": 0.00564 |
| }, |
| { |
| "epoch": 0.2487852283770651, |
| "grad_norm": 2.2867326736450195, |
| "learning_rate": 0.00019941576092482102, |
| "logits/chosen": -0.7742420434951782, |
| "logits/rejected": -0.7834679484367371, |
| "logps/chosen": -16.7891788482666, |
| "logps/rejected": -41.97465133666992, |
| "loss": 0.8852651119232178, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5684996843338013, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.13892318308353424, |
| "rewards/margins": 1.5296766757965088, |
| "rewards/rejected": -1.3907533884048462, |
| "step": 104, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.25117739403453687, |
| "grad_norm": 2.7168843746185303, |
| "learning_rate": 0.00019938694341337397, |
| "logits/chosen": -0.761275053024292, |
| "logits/rejected": -0.7697631120681763, |
| "logps/chosen": -7.466435432434082, |
| "logps/rejected": -35.15258026123047, |
| "loss": 1.0209513902664185, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.636327862739563, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.23773828148841858, |
| "rewards/margins": 1.3149192333221436, |
| "rewards/rejected": -1.0771809816360474, |
| "step": 105, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.2535695596920087, |
| "grad_norm": 3.2692759037017822, |
| "learning_rate": 0.00019935743438103564, |
| "logits/chosen": -0.7422237992286682, |
| "logits/rejected": -0.7514990568161011, |
| "logps/chosen": -10.624504089355469, |
| "logps/rejected": -45.264095306396484, |
| "loss": 0.860304057598114, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5044058561325073, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.3707730770111084, |
| "rewards/margins": 1.6010938882827759, |
| "rewards/rejected": -1.230320930480957, |
| "step": 106, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.2559617253494805, |
| "grad_norm": 2.6627414226531982, |
| "learning_rate": 0.0001993272340331259, |
| "logits/chosen": -0.75310879945755, |
| "logits/rejected": -0.7589280605316162, |
| "logps/chosen": -11.713288307189941, |
| "logps/rejected": -42.20861053466797, |
| "loss": 0.7711183428764343, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.43505772948265076, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.3599666357040405, |
| "rewards/margins": 1.9370594024658203, |
| "rewards/rejected": -1.5770928859710693, |
| "step": 107, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.2583538910069522, |
| "grad_norm": 10.524235725402832, |
| "learning_rate": 0.00019929634257977467, |
| "logits/chosen": -0.7487523555755615, |
| "logits/rejected": -0.7474150657653809, |
| "logps/chosen": -13.211384773254395, |
| "logps/rejected": -29.434967041015625, |
| "loss": 1.4064631462097168, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.9639731049537659, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.05272696912288666, |
| "rewards/margins": 1.349626064300537, |
| "rewards/rejected": -1.2968993186950684, |
| "step": 108, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.260746056664424, |
| "grad_norm": 3.4797213077545166, |
| "learning_rate": 0.00019926476023592052, |
| "logits/chosen": -0.7358635067939758, |
| "logits/rejected": -0.7399594783782959, |
| "logps/chosen": -10.68916130065918, |
| "logps/rejected": -39.55656814575195, |
| "loss": 0.8483824133872986, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.512445867061615, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.26777422428131104, |
| "rewards/margins": 1.7832790613174438, |
| "rewards/rejected": -1.5155047178268433, |
| "step": 109, |
| "train_speed(iter/s)": 0.00564 |
| }, |
| { |
| "epoch": 0.26313822232189576, |
| "grad_norm": 5.318943023681641, |
| "learning_rate": 0.00019923248722130907, |
| "logits/chosen": -0.8071874976158142, |
| "logits/rejected": -0.8125062584877014, |
| "logps/chosen": -13.162405967712402, |
| "logps/rejected": -41.4532356262207, |
| "loss": 0.6827958822250366, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.39053356647491455, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.33915120363235474, |
| "rewards/margins": 1.944946050643921, |
| "rewards/rejected": -1.605794906616211, |
| "step": 110, |
| "train_speed(iter/s)": 0.00564 |
| }, |
| { |
| "epoch": 0.26553038797936757, |
| "grad_norm": 6.246517658233643, |
| "learning_rate": 0.0001991995237604916, |
| "logits/chosen": -0.7993965148925781, |
| "logits/rejected": -0.8035753965377808, |
| "logps/chosen": -7.477054595947266, |
| "logps/rejected": -43.46379089355469, |
| "loss": 0.6327071785926819, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.3665684461593628, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.18014857172966003, |
| "rewards/margins": 2.2636728286743164, |
| "rewards/rejected": -2.083524227142334, |
| "step": 111, |
| "train_speed(iter/s)": 0.005641 |
| }, |
| { |
| "epoch": 0.26792255363683937, |
| "grad_norm": 2.495997905731201, |
| "learning_rate": 0.00019916587008282345, |
| "logits/chosen": -0.8536009192466736, |
| "logits/rejected": -0.8555616736412048, |
| "logps/chosen": -12.577634811401367, |
| "logps/rejected": -40.71754837036133, |
| "loss": 0.7153143286705017, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4246208667755127, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.16410964727401733, |
| "rewards/margins": 2.231029748916626, |
| "rewards/rejected": -2.066920042037964, |
| "step": 112, |
| "train_speed(iter/s)": 0.005641 |
| }, |
| { |
| "epoch": 0.2703147192943111, |
| "grad_norm": 2.619420289993286, |
| "learning_rate": 0.00019913152642246233, |
| "logits/chosen": -0.8602064847946167, |
| "logits/rejected": -0.870837926864624, |
| "logps/chosen": -10.31131362915039, |
| "logps/rejected": -69.16264343261719, |
| "loss": 0.6141177415847778, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4619813561439514, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.4243915379047394, |
| "rewards/margins": 3.666055202484131, |
| "rewards/rejected": -3.241663694381714, |
| "step": 113, |
| "train_speed(iter/s)": 0.00564 |
| }, |
| { |
| "epoch": 0.2727068849517829, |
| "grad_norm": 3.2425930500030518, |
| "learning_rate": 0.00019909649301836675, |
| "logits/chosen": -0.8935564756393433, |
| "logits/rejected": -0.8967443108558655, |
| "logps/chosen": -11.790775299072266, |
| "logps/rejected": -54.12834930419922, |
| "loss": 0.6750085353851318, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.44180190563201904, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.022756237536668777, |
| "rewards/margins": 2.9651176929473877, |
| "rewards/rejected": -2.942361354827881, |
| "step": 114, |
| "train_speed(iter/s)": 0.005639 |
| }, |
| { |
| "epoch": 0.2750990506092547, |
| "grad_norm": 24.632802963256836, |
| "learning_rate": 0.0001990607701142944, |
| "logits/chosen": -0.8810749650001526, |
| "logits/rejected": -0.8888768553733826, |
| "logps/chosen": -23.208826065063477, |
| "logps/rejected": -53.65020751953125, |
| "loss": 1.851314902305603, |
| "memory(GiB)": 46.78, |
| "nll_loss": 1.315173625946045, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.5939430594444275, |
| "rewards/margins": 2.5491535663604736, |
| "rewards/rejected": -3.143096685409546, |
| "step": 115, |
| "train_speed(iter/s)": 0.005639 |
| }, |
| { |
| "epoch": 0.27749121626672646, |
| "grad_norm": 17.350990295410156, |
| "learning_rate": 0.00019902435795880044, |
| "logits/chosen": -0.9093245267868042, |
| "logits/rejected": -0.9126628637313843, |
| "logps/chosen": -16.813339233398438, |
| "logps/rejected": -48.75463104248047, |
| "loss": 1.4077650308609009, |
| "memory(GiB)": 46.78, |
| "nll_loss": 1.050284743309021, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.2693726122379303, |
| "rewards/margins": 2.530968427658081, |
| "rewards/rejected": -2.8003406524658203, |
| "step": 116, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.27988338192419826, |
| "grad_norm": 4.22681188583374, |
| "learning_rate": 0.00019898725680523568, |
| "logits/chosen": -0.8980717062950134, |
| "logits/rejected": -0.9024423956871033, |
| "logps/chosen": -17.1817626953125, |
| "logps/rejected": -39.5125732421875, |
| "loss": 1.0394302606582642, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6766522526741028, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.015612306073307991, |
| "rewards/margins": 2.2480926513671875, |
| "rewards/rejected": -2.23248028755188, |
| "step": 117, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.28227554758167, |
| "grad_norm": 4.9705047607421875, |
| "learning_rate": 0.00019894946691174494, |
| "logits/chosen": -0.9122129082679749, |
| "logits/rejected": -0.9150456190109253, |
| "logps/chosen": -14.797954559326172, |
| "logps/rejected": -38.31578826904297, |
| "loss": 1.0148959159851074, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6429183483123779, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.06372372806072235, |
| "rewards/margins": 2.232508659362793, |
| "rewards/rejected": -2.296232223510742, |
| "step": 118, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.2846677132391418, |
| "grad_norm": 5.455178737640381, |
| "learning_rate": 0.00019891098854126518, |
| "logits/chosen": -0.8895492553710938, |
| "logits/rejected": -0.8904259204864502, |
| "logps/chosen": -15.091131210327148, |
| "logps/rejected": -40.248023986816406, |
| "loss": 1.0479692220687866, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5974382758140564, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.1267915964126587, |
| "rewards/margins": 1.8017388582229614, |
| "rewards/rejected": -1.9285303354263306, |
| "step": 119, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.2870598788966136, |
| "grad_norm": 3.6146156787872314, |
| "learning_rate": 0.00019887182196152368, |
| "logits/chosen": -0.8824920654296875, |
| "logits/rejected": -0.8917353749275208, |
| "logps/chosen": -7.259885311126709, |
| "logps/rejected": -55.807106018066406, |
| "loss": 0.6745078563690186, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.3909413814544678, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12902025878429413, |
| "rewards/margins": 2.945380210876465, |
| "rewards/rejected": -2.8163604736328125, |
| "step": 120, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.28945204455408535, |
| "grad_norm": 2.7159321308135986, |
| "learning_rate": 0.0001988319674450362, |
| "logits/chosen": -0.8625448942184448, |
| "logits/rejected": -0.8625946640968323, |
| "logps/chosen": -14.481998443603516, |
| "logps/rejected": -32.343990325927734, |
| "loss": 0.8893653750419617, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5466328859329224, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1943284571170807, |
| "rewards/margins": 1.5906785726547241, |
| "rewards/rejected": -1.3963501453399658, |
| "step": 121, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.29184421021155715, |
| "grad_norm": 3.480830430984497, |
| "learning_rate": 0.0001987914252691051, |
| "logits/chosen": -0.854420006275177, |
| "logits/rejected": -0.8605158925056458, |
| "logps/chosen": -11.787839889526367, |
| "logps/rejected": -43.2979850769043, |
| "loss": 0.7196720838546753, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5012603402137756, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2540440559387207, |
| "rewards/margins": 2.362658977508545, |
| "rewards/rejected": -2.1086151599884033, |
| "step": 122, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.29423637586902895, |
| "grad_norm": 3.526841163635254, |
| "learning_rate": 0.00019875019571581728, |
| "logits/chosen": -0.8358349204063416, |
| "logits/rejected": -0.8408463597297668, |
| "logps/chosen": -10.75222110748291, |
| "logps/rejected": -39.87123489379883, |
| "loss": 0.8948889970779419, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5788209438323975, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.15971140563488007, |
| "rewards/margins": 2.309971809387207, |
| "rewards/rejected": -2.1502604484558105, |
| "step": 123, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.2966285415265007, |
| "grad_norm": 2.6714415550231934, |
| "learning_rate": 0.00019870827907204243, |
| "logits/chosen": -0.8233751654624939, |
| "logits/rejected": -0.8268409967422485, |
| "logps/chosen": -7.600090503692627, |
| "logps/rejected": -23.002704620361328, |
| "loss": 0.8616309762001038, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.43116894364356995, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.28889453411102295, |
| "rewards/margins": 1.081323504447937, |
| "rewards/rejected": -0.7924288511276245, |
| "step": 124, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.2990207071839725, |
| "grad_norm": 2.5394530296325684, |
| "learning_rate": 0.0001986656756294309, |
| "logits/chosen": -0.8398711085319519, |
| "logits/rejected": -0.8451254963874817, |
| "logps/chosen": -9.005830764770508, |
| "logps/rejected": -33.04253387451172, |
| "loss": 0.8945890665054321, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4701317250728607, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.3293447494506836, |
| "rewards/margins": 1.3195619583129883, |
| "rewards/rejected": -0.9902173280715942, |
| "step": 125, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.30141287284144425, |
| "grad_norm": 2.8723111152648926, |
| "learning_rate": 0.00019862238568441165, |
| "logits/chosen": -0.815192699432373, |
| "logits/rejected": -0.8217712044715881, |
| "logps/chosen": -9.070804595947266, |
| "logps/rejected": -33.22805404663086, |
| "loss": 0.7637823820114136, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.38797256350517273, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.17756348848342896, |
| "rewards/margins": 1.2646557092666626, |
| "rewards/rejected": -1.0870922803878784, |
| "step": 126, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.30380503849891605, |
| "grad_norm": 2.79227614402771, |
| "learning_rate": 0.00019857840953819026, |
| "logits/chosen": -0.8166335821151733, |
| "logits/rejected": -0.8273878693580627, |
| "logps/chosen": -9.302844047546387, |
| "logps/rejected": -32.995567321777344, |
| "loss": 0.7958816289901733, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4939688444137573, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.3217732310295105, |
| "rewards/margins": 1.6436134576797485, |
| "rewards/rejected": -1.3218401670455933, |
| "step": 127, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.30619720415638785, |
| "grad_norm": 2.4456770420074463, |
| "learning_rate": 0.00019853374749674677, |
| "logits/chosen": -0.8173456192016602, |
| "logits/rejected": -0.8242702484130859, |
| "logps/chosen": -10.526348114013672, |
| "logps/rejected": -49.73170471191406, |
| "loss": 0.6932899951934814, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.410495400428772, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2942225933074951, |
| "rewards/margins": 2.532844305038452, |
| "rewards/rejected": -2.238621711730957, |
| "step": 128, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.3085893698138596, |
| "grad_norm": 9.124894142150879, |
| "learning_rate": 0.00019848839987083366, |
| "logits/chosen": -0.8869707584381104, |
| "logits/rejected": -0.8904052376747131, |
| "logps/chosen": -15.062342643737793, |
| "logps/rejected": -45.34114456176758, |
| "loss": 1.0228577852249146, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5322936773300171, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.14518728852272034, |
| "rewards/margins": 2.259458541870117, |
| "rewards/rejected": -2.4046459197998047, |
| "step": 129, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.3109815354713314, |
| "grad_norm": 3.8776865005493164, |
| "learning_rate": 0.00019844236697597354, |
| "logits/chosen": -0.8990283012390137, |
| "logits/rejected": -0.912631630897522, |
| "logps/chosen": -9.924826622009277, |
| "logps/rejected": -59.806114196777344, |
| "loss": 0.6793578863143921, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4696520268917084, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.21068614721298218, |
| "rewards/margins": 3.6280405521392822, |
| "rewards/rejected": -3.4173545837402344, |
| "step": 130, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.3133737011288032, |
| "grad_norm": 5.517581462860107, |
| "learning_rate": 0.0001983956491324571, |
| "logits/chosen": -0.890454888343811, |
| "logits/rejected": -0.8954933881759644, |
| "logps/chosen": -11.391611099243164, |
| "logps/rejected": -50.16078567504883, |
| "loss": 0.9601025581359863, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.588812530040741, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.01075834035873413, |
| "rewards/margins": 3.3035926818847656, |
| "rewards/rejected": -3.292834758758545, |
| "step": 131, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.31576586678627494, |
| "grad_norm": 2.991635322570801, |
| "learning_rate": 0.0001983482466653407, |
| "logits/chosen": -0.8738813400268555, |
| "logits/rejected": -0.870830774307251, |
| "logps/chosen": -8.894745826721191, |
| "logps/rejected": -39.2451171875, |
| "loss": 0.6562129855155945, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.34973058104515076, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2841479778289795, |
| "rewards/margins": 2.604597330093384, |
| "rewards/rejected": -2.3204493522644043, |
| "step": 132, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.31815803244374674, |
| "grad_norm": 8.708478927612305, |
| "learning_rate": 0.00019830015990444434, |
| "logits/chosen": -0.8821411728858948, |
| "logits/rejected": -0.891589879989624, |
| "logps/chosen": -18.36916732788086, |
| "logps/rejected": -58.93795394897461, |
| "loss": 0.8125941753387451, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6013258695602417, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.01710733026266098, |
| "rewards/margins": 2.8880860805511475, |
| "rewards/rejected": -2.905193567276001, |
| "step": 133, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.3205501981012185, |
| "grad_norm": 5.8355326652526855, |
| "learning_rate": 0.00019825138918434915, |
| "logits/chosen": -0.8907016515731812, |
| "logits/rejected": -0.9023689031600952, |
| "logps/chosen": -7.357381343841553, |
| "logps/rejected": -43.13167190551758, |
| "loss": 0.8315655589103699, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5653954148292542, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.18780827522277832, |
| "rewards/margins": 2.5148556232452393, |
| "rewards/rejected": -2.327047348022461, |
| "step": 134, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.3229423637586903, |
| "grad_norm": 5.084504127502441, |
| "learning_rate": 0.0001982019348443952, |
| "logits/chosen": -0.8744340538978577, |
| "logits/rejected": -0.8839950561523438, |
| "logps/chosen": -11.753833770751953, |
| "logps/rejected": -50.23076248168945, |
| "loss": 0.900967001914978, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6376667618751526, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2352873831987381, |
| "rewards/margins": 2.9518003463745117, |
| "rewards/rejected": -2.71651291847229, |
| "step": 135, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.3253345294161621, |
| "grad_norm": 7.443116664886475, |
| "learning_rate": 0.00019815179722867913, |
| "logits/chosen": -0.8482657670974731, |
| "logits/rejected": -0.8486264944076538, |
| "logps/chosen": -13.961491584777832, |
| "logps/rejected": -33.860713958740234, |
| "loss": 1.0923020839691162, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.626233696937561, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.035439833998680115, |
| "rewards/margins": 1.6097826957702637, |
| "rewards/rejected": -1.574342966079712, |
| "step": 136, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.32772669507363383, |
| "grad_norm": 6.069121360778809, |
| "learning_rate": 0.0001981009766860517, |
| "logits/chosen": -0.8712698817253113, |
| "logits/rejected": -0.8716377019882202, |
| "logps/chosen": -12.955726623535156, |
| "logps/rejected": -39.7027587890625, |
| "loss": 1.063010811805725, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.7573559284210205, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.1307298243045807, |
| "rewards/margins": 2.164278507232666, |
| "rewards/rejected": -2.295008420944214, |
| "step": 137, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.33011886073110563, |
| "grad_norm": 4.233928203582764, |
| "learning_rate": 0.00019804947357011522, |
| "logits/chosen": -0.8236069679260254, |
| "logits/rejected": -0.834182858467102, |
| "logps/chosen": -11.049253463745117, |
| "logps/rejected": -44.08565902709961, |
| "loss": 0.8465083241462708, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5441975593566895, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.11046641319990158, |
| "rewards/margins": 2.231780767440796, |
| "rewards/rejected": -2.121314525604248, |
| "step": 138, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.33251102638857744, |
| "grad_norm": 4.2061309814453125, |
| "learning_rate": 0.00019799728823922155, |
| "logits/chosen": -0.8435903191566467, |
| "logits/rejected": -0.8470354080200195, |
| "logps/chosen": -8.918343544006348, |
| "logps/rejected": -36.74750900268555, |
| "loss": 0.7144109010696411, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.40702298283576965, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.38820797204971313, |
| "rewards/margins": 1.9121365547180176, |
| "rewards/rejected": -1.5239286422729492, |
| "step": 139, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.3349031920460492, |
| "grad_norm": 4.416022300720215, |
| "learning_rate": 0.0001979444210564691, |
| "logits/chosen": -0.7840769290924072, |
| "logits/rejected": -0.7934098839759827, |
| "logps/chosen": -8.80458927154541, |
| "logps/rejected": -47.96394348144531, |
| "loss": 0.6305175423622131, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.46931201219558716, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.1097404956817627, |
| "rewards/margins": 2.367628335952759, |
| "rewards/rejected": -2.257887840270996, |
| "step": 140, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.337295357703521, |
| "grad_norm": 4.785271644592285, |
| "learning_rate": 0.00019789087238970054, |
| "logits/chosen": -0.8002479076385498, |
| "logits/rejected": -0.8085525035858154, |
| "logps/chosen": -13.70367431640625, |
| "logps/rejected": -50.61127853393555, |
| "loss": 0.9849874377250671, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6505732536315918, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.002065986394882202, |
| "rewards/margins": 2.2779290676116943, |
| "rewards/rejected": -2.2758634090423584, |
| "step": 141, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.3396875233609927, |
| "grad_norm": 4.302982330322266, |
| "learning_rate": 0.0001978366426115003, |
| "logits/chosen": -0.7700361013412476, |
| "logits/rejected": -0.7705273628234863, |
| "logps/chosen": -7.580388069152832, |
| "logps/rejected": -34.4312744140625, |
| "loss": 0.5671791434288025, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.33011525869369507, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.3449672758579254, |
| "rewards/margins": 2.0864133834838867, |
| "rewards/rejected": -1.7414460182189941, |
| "step": 142, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.34207968901846453, |
| "grad_norm": 4.768328666687012, |
| "learning_rate": 0.00019778173209919174, |
| "logits/chosen": -0.7594372034072876, |
| "logits/rejected": -0.7637089490890503, |
| "logps/chosen": -14.413496971130371, |
| "logps/rejected": -35.399288177490234, |
| "loss": 0.8721976280212402, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6180708408355713, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.4894520044326782, |
| "rewards/margins": 1.8526508808135986, |
| "rewards/rejected": -1.3631988763809204, |
| "step": 143, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.34447185467593633, |
| "grad_norm": 2.74961519241333, |
| "learning_rate": 0.00019772614123483487, |
| "logits/chosen": -0.778110146522522, |
| "logits/rejected": -0.7844111919403076, |
| "logps/chosen": -15.620856285095215, |
| "logps/rejected": -41.484710693359375, |
| "loss": 0.7825542092323303, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.49021872878074646, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1902773678302765, |
| "rewards/margins": 1.998608112335205, |
| "rewards/rejected": -1.808330774307251, |
| "step": 144, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.3468640203334081, |
| "grad_norm": 4.053314208984375, |
| "learning_rate": 0.00019766987040522333, |
| "logits/chosen": -0.8054400682449341, |
| "logits/rejected": -0.8061996698379517, |
| "logps/chosen": -11.948256492614746, |
| "logps/rejected": -25.110414505004883, |
| "loss": 1.033211350440979, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.6106947064399719, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.13057959079742432, |
| "rewards/margins": 1.2587168216705322, |
| "rewards/rejected": -1.128137230873108, |
| "step": 145, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.3492561859908799, |
| "grad_norm": 2.9867749214172363, |
| "learning_rate": 0.00019761292000188197, |
| "logits/chosen": -0.8127353191375732, |
| "logits/rejected": -0.8201886415481567, |
| "logps/chosen": -16.549510955810547, |
| "logps/rejected": -38.11656188964844, |
| "loss": 0.9068831205368042, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.5614981055259705, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.20250406861305237, |
| "rewards/margins": 1.7395099401474, |
| "rewards/rejected": -1.5370057821273804, |
| "step": 146, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.3516483516483517, |
| "grad_norm": 2.311192750930786, |
| "learning_rate": 0.00019755529042106394, |
| "logits/chosen": -0.7973526120185852, |
| "logits/rejected": -0.800200343132019, |
| "logps/chosen": -7.481082916259766, |
| "logps/rejected": -31.512645721435547, |
| "loss": 0.7267748713493347, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.36258465051651, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.21067243814468384, |
| "rewards/margins": 1.607694149017334, |
| "rewards/rejected": -1.3970215320587158, |
| "step": 147, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.3540405173058234, |
| "grad_norm": 2.8598074913024902, |
| "learning_rate": 0.00019749698206374808, |
| "logits/chosen": -0.849799633026123, |
| "logits/rejected": -0.8563861846923828, |
| "logps/chosen": -9.258435249328613, |
| "logps/rejected": -43.992244720458984, |
| "loss": 0.6600675582885742, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4354017972946167, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.33977681398391724, |
| "rewards/margins": 2.25696063041687, |
| "rewards/rejected": -1.9171836376190186, |
| "step": 148, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.3564326829632952, |
| "grad_norm": 2.3342740535736084, |
| "learning_rate": 0.00019743799533563603, |
| "logits/chosen": -0.8605116605758667, |
| "logits/rejected": -0.8653464913368225, |
| "logps/chosen": -7.782684326171875, |
| "logps/rejected": -37.683738708496094, |
| "loss": 0.6275639533996582, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.4273594915866852, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.4285268187522888, |
| "rewards/margins": 2.251387357711792, |
| "rewards/rejected": -1.8228604793548584, |
| "step": 149, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.35882484862076697, |
| "grad_norm": 4.0247483253479, |
| "learning_rate": 0.0001973783306471495, |
| "logits/chosen": -0.8887797594070435, |
| "logits/rejected": -0.898193895816803, |
| "logps/chosen": -8.930159568786621, |
| "logps/rejected": -43.03559112548828, |
| "loss": 0.7508974671363831, |
| "memory(GiB)": 46.78, |
| "nll_loss": 0.518282413482666, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.28456175327301025, |
| "rewards/margins": 2.084869146347046, |
| "rewards/rejected": -1.8003073930740356, |
| "step": 150, |
| "train_speed(iter/s)": 0.005634 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 1254, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "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": 1.0456400650462822e+18, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
