| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 775, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012903225806451613, | |
| "grad_norm": 21.62714956675265, | |
| "learning_rate": 9.000000000000001e-07, | |
| "log_odds_chosen": 1.0578124523162842, | |
| "log_odds_ratio": -0.3404296934604645, | |
| "logits/chosen": -0.9966796636581421, | |
| "logits/rejected": 15.725000381469727, | |
| "logps/chosen": -1.642187476158142, | |
| "logps/rejected": -2.565624952316284, | |
| "loss": 1.9877, | |
| "nll_loss": 1.970312476158142, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.08212890475988388, | |
| "rewards/margins": 0.04628906399011612, | |
| "rewards/rejected": -0.12832030653953552, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.025806451612903226, | |
| "grad_norm": 9.886558420099602, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "log_odds_chosen": 0.8812500238418579, | |
| "log_odds_ratio": -0.38496094942092896, | |
| "logits/chosen": -0.22133788466453552, | |
| "logits/rejected": 17.149999618530273, | |
| "logps/chosen": -1.5164062976837158, | |
| "logps/rejected": -2.260937452316284, | |
| "loss": 1.7895, | |
| "nll_loss": 1.76953125, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.07583007961511612, | |
| "rewards/margins": 0.03732910007238388, | |
| "rewards/rejected": -0.11313476413488388, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03870967741935484, | |
| "grad_norm": 7.174101275947782, | |
| "learning_rate": 2.9e-06, | |
| "log_odds_chosen": 0.49199217557907104, | |
| "log_odds_ratio": -0.508984386920929, | |
| "logits/chosen": 0.4482788145542145, | |
| "logits/rejected": 17.043750762939453, | |
| "logps/chosen": -1.3585937023162842, | |
| "logps/rejected": -1.7414062023162842, | |
| "loss": 1.5844, | |
| "nll_loss": 1.55859375, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.06796874850988388, | |
| "rewards/margins": 0.01908569410443306, | |
| "rewards/rejected": -0.08696289360523224, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05161290322580645, | |
| "grad_norm": 5.02065428217483, | |
| "learning_rate": 3.900000000000001e-06, | |
| "log_odds_chosen": 0.3095703125, | |
| "log_odds_ratio": -0.5765625238418579, | |
| "logits/chosen": 1.201074242591858, | |
| "logits/rejected": 17.387500762939453, | |
| "logps/chosen": -1.1945312023162842, | |
| "logps/rejected": -1.423437476158142, | |
| "loss": 1.3805, | |
| "nll_loss": 1.3515625, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.05979003757238388, | |
| "rewards/margins": 0.011395263485610485, | |
| "rewards/rejected": -0.07119140774011612, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 4.574739753680098, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "log_odds_chosen": 0.24362793564796448, | |
| "log_odds_ratio": -0.614453136920929, | |
| "logits/chosen": 0.5683349370956421, | |
| "logits/rejected": 17.09375, | |
| "logps/chosen": -1.123437523841858, | |
| "logps/rejected": -1.303125023841858, | |
| "loss": 1.2454, | |
| "nll_loss": 1.21484375, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.05605468899011612, | |
| "rewards/margins": 0.00897369347512722, | |
| "rewards/rejected": -0.0650634765625, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07741935483870968, | |
| "grad_norm": 5.580591326486056, | |
| "learning_rate": 5.9e-06, | |
| "log_odds_chosen": 0.2967773377895355, | |
| "log_odds_ratio": -0.5923827886581421, | |
| "logits/chosen": 0.01669921912252903, | |
| "logits/rejected": 16.193750381469727, | |
| "logps/chosen": -1.1171875, | |
| "logps/rejected": -1.3359375, | |
| "loss": 1.2094, | |
| "nll_loss": 1.1796875, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.05585937574505806, | |
| "rewards/margins": 0.01091918908059597, | |
| "rewards/rejected": -0.06684570014476776, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09032258064516129, | |
| "grad_norm": 4.465129508883455, | |
| "learning_rate": 6.9e-06, | |
| "log_odds_chosen": 0.928906261920929, | |
| "log_odds_ratio": -0.3726562559604645, | |
| "logits/chosen": -1.2811768054962158, | |
| "logits/rejected": 15.90625, | |
| "logps/chosen": -1.033593773841858, | |
| "logps/rejected": -1.740625023841858, | |
| "loss": 1.0955, | |
| "nll_loss": 1.0773437023162842, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.0516357421875, | |
| "rewards/margins": 0.03531493991613388, | |
| "rewards/rejected": -0.08698730170726776, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1032258064516129, | |
| "grad_norm": 4.325447268788425, | |
| "learning_rate": 7.9e-06, | |
| "log_odds_chosen": 1.450781226158142, | |
| "log_odds_ratio": -0.2611328065395355, | |
| "logits/chosen": -1.3720703125, | |
| "logits/rejected": 15.731249809265137, | |
| "logps/chosen": -1.071874976158142, | |
| "logps/rejected": -2.2242188453674316, | |
| "loss": 1.0965, | |
| "nll_loss": 1.0828125476837158, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.0535888671875, | |
| "rewards/margins": 0.0576171875, | |
| "rewards/rejected": -0.11123047024011612, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11612903225806452, | |
| "grad_norm": 4.408329645746575, | |
| "learning_rate": 8.900000000000001e-06, | |
| "log_odds_chosen": 1.787500023841858, | |
| "log_odds_ratio": -0.23085936903953552, | |
| "logits/chosen": -1.55859375, | |
| "logits/rejected": 15.662500381469727, | |
| "logps/chosen": -1.04296875, | |
| "logps/rejected": -2.499218702316284, | |
| "loss": 1.0693, | |
| "nll_loss": 1.057031273841858, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.0521240234375, | |
| "rewards/margins": 0.07275390625, | |
| "rewards/rejected": -0.12495116889476776, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 5.360425933748804, | |
| "learning_rate": 9.9e-06, | |
| "log_odds_chosen": 1.853906273841858, | |
| "log_odds_ratio": -0.21064452826976776, | |
| "logits/chosen": -1.640234351158142, | |
| "logits/rejected": 15.243749618530273, | |
| "logps/chosen": -1.072656273841858, | |
| "logps/rejected": -2.6078124046325684, | |
| "loss": 1.1013, | |
| "nll_loss": 1.091406226158142, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.05366211012005806, | |
| "rewards/margins": 0.07661132514476776, | |
| "rewards/rejected": -0.13032226264476776, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14193548387096774, | |
| "grad_norm": 4.490699979708051, | |
| "learning_rate": 9.578262852211514e-06, | |
| "log_odds_chosen": 2.003124952316284, | |
| "log_odds_ratio": -0.19697265326976776, | |
| "logits/chosen": -1.007055640220642, | |
| "logits/rejected": 16.012500762939453, | |
| "logps/chosen": -1.056640625, | |
| "logps/rejected": -2.7171874046325684, | |
| "loss": 1.0846, | |
| "nll_loss": 1.0753905773162842, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.05283202975988388, | |
| "rewards/margins": 0.08295898139476776, | |
| "rewards/rejected": -0.13593749701976776, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15483870967741936, | |
| "grad_norm": 4.549502496283154, | |
| "learning_rate": 9.166984970282113e-06, | |
| "log_odds_chosen": 2.311718702316284, | |
| "log_odds_ratio": -0.16552734375, | |
| "logits/chosen": -1.816796898841858, | |
| "logits/rejected": 16.3125, | |
| "logps/chosen": -1.0390625, | |
| "logps/rejected": -2.973437547683716, | |
| "loss": 1.058, | |
| "nll_loss": 1.05078125, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.05197753757238388, | |
| "rewards/margins": 0.09672851860523224, | |
| "rewards/rejected": -0.14873047173023224, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16774193548387098, | |
| "grad_norm": 4.492595553477561, | |
| "learning_rate": 8.804509063256239e-06, | |
| "log_odds_chosen": 2.5640625953674316, | |
| "log_odds_ratio": -0.13984374701976776, | |
| "logits/chosen": -1.11279296875, | |
| "logits/rejected": 15.949999809265137, | |
| "logps/chosen": -1.0703125, | |
| "logps/rejected": -3.253124952316284, | |
| "loss": 1.0883, | |
| "nll_loss": 1.0812499523162842, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.0535888671875, | |
| "rewards/margins": 0.10903320461511612, | |
| "rewards/rejected": -0.16269531846046448, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18064516129032257, | |
| "grad_norm": 4.778523731144206, | |
| "learning_rate": 8.481889296799709e-06, | |
| "log_odds_chosen": 2.4632811546325684, | |
| "log_odds_ratio": -0.14370116591453552, | |
| "logits/chosen": -1.4695312976837158, | |
| "logits/rejected": 15.431249618530273, | |
| "logps/chosen": -1.04296875, | |
| "logps/rejected": -3.1171875, | |
| "loss": 1.0656, | |
| "nll_loss": 1.0574219226837158, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05217285081744194, | |
| "rewards/margins": 0.103759765625, | |
| "rewards/rejected": -0.15595702826976776, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 4.292339667379623, | |
| "learning_rate": 8.192319205190406e-06, | |
| "log_odds_chosen": 2.5328125953674316, | |
| "log_odds_ratio": -0.13725586235523224, | |
| "logits/chosen": -2.3530516624450684, | |
| "logits/rejected": 15.293749809265137, | |
| "logps/chosen": -1.0382812023162842, | |
| "logps/rejected": -3.1812500953674316, | |
| "loss": 1.0556, | |
| "nll_loss": 1.049218773841858, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.0518798828125, | |
| "rewards/margins": 0.10727538913488388, | |
| "rewards/rejected": -0.15888671576976776, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2064516129032258, | |
| "grad_norm": 5.246753687109851, | |
| "learning_rate": 7.930515857181441e-06, | |
| "log_odds_chosen": 2.317187547683716, | |
| "log_odds_ratio": -0.1767578125, | |
| "logits/chosen": -2.588671922683716, | |
| "logits/rejected": 14.71875, | |
| "logps/chosen": -1.023828148841858, | |
| "logps/rejected": -2.965625047683716, | |
| "loss": 1.0458, | |
| "nll_loss": 1.0363280773162842, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05119628831744194, | |
| "rewards/margins": 0.09697265923023224, | |
| "rewards/rejected": -0.14814452826976776, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21935483870967742, | |
| "grad_norm": 4.346066733666396, | |
| "learning_rate": 7.692307692307692e-06, | |
| "log_odds_chosen": 2.692187547683716, | |
| "log_odds_ratio": -0.13859863579273224, | |
| "logits/chosen": -2.8062500953674316, | |
| "logits/rejected": 15.743749618530273, | |
| "logps/chosen": -0.9671875238418579, | |
| "logps/rejected": -3.2515625953674316, | |
| "loss": 0.9999, | |
| "nll_loss": 0.9917968511581421, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04836425930261612, | |
| "rewards/margins": 0.11416015774011612, | |
| "rewards/rejected": -0.16249999403953552, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.23225806451612904, | |
| "grad_norm": 4.703610048799506, | |
| "learning_rate": 7.4743509275193595e-06, | |
| "log_odds_chosen": 2.8375000953674316, | |
| "log_odds_ratio": -0.12353515625, | |
| "logits/chosen": -2.365234375, | |
| "logits/rejected": 15.206250190734863, | |
| "logps/chosen": -0.9917968511581421, | |
| "logps/rejected": -3.410937547683716, | |
| "loss": 1.0064, | |
| "nll_loss": 1.0007812976837158, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04960937425494194, | |
| "rewards/margins": 0.12089844048023224, | |
| "rewards/rejected": -0.17060546576976776, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24516129032258063, | |
| "grad_norm": 4.322341540802487, | |
| "learning_rate": 7.27392967453308e-06, | |
| "log_odds_chosen": 2.792187452316284, | |
| "log_odds_ratio": -0.13710936903953552, | |
| "logits/chosen": -1.894384741783142, | |
| "logits/rejected": 15.631250381469727, | |
| "logps/chosen": -0.9921875, | |
| "logps/rejected": -3.375, | |
| "loss": 1.0165, | |
| "nll_loss": 1.010156273841858, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.049560546875, | |
| "rewards/margins": 0.119140625, | |
| "rewards/rejected": -0.16884765028953552, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 4.113404991486903, | |
| "learning_rate": 7.088812050083359e-06, | |
| "log_odds_chosen": 3.575000047683716, | |
| "log_odds_ratio": -0.0809326171875, | |
| "logits/chosen": -2.4609375, | |
| "logits/rejected": 15.800000190734863, | |
| "logps/chosen": -1.013671875, | |
| "logps/rejected": -4.154687404632568, | |
| "loss": 1.0316, | |
| "nll_loss": 1.0281250476837158, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05070800706744194, | |
| "rewards/margins": 0.15688475966453552, | |
| "rewards/rejected": -0.20761719346046448, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2709677419354839, | |
| "grad_norm": 4.585624983468261, | |
| "learning_rate": 6.917144638660746e-06, | |
| "log_odds_chosen": 3.4140625, | |
| "log_odds_ratio": -0.12514647841453552, | |
| "logits/chosen": -2.740234375, | |
| "logits/rejected": 15.024999618530273, | |
| "logps/chosen": -1.008203148841858, | |
| "logps/rejected": -4.001562595367432, | |
| "loss": 1.0165, | |
| "nll_loss": 1.009765625, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.0504150390625, | |
| "rewards/margins": 0.149658203125, | |
| "rewards/rejected": -0.20000000298023224, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2838709677419355, | |
| "grad_norm": 4.454926501194002, | |
| "learning_rate": 6.757373783994861e-06, | |
| "log_odds_chosen": 2.43359375, | |
| "log_odds_ratio": -0.177978515625, | |
| "logits/chosen": -1.859765648841858, | |
| "logits/rejected": 15.618749618530273, | |
| "logps/chosen": -1.05859375, | |
| "logps/rejected": -3.1265625953674316, | |
| "loss": 1.0732, | |
| "nll_loss": 1.0656249523162842, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.05292968824505806, | |
| "rewards/margins": 0.103515625, | |
| "rewards/rejected": -0.15634766221046448, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2967741935483871, | |
| "grad_norm": 4.239789629191708, | |
| "learning_rate": 6.6081860045508985e-06, | |
| "log_odds_chosen": 2.297656297683716, | |
| "log_odds_ratio": -0.173583984375, | |
| "logits/chosen": -1.5969727039337158, | |
| "logits/rejected": 15.318750381469727, | |
| "logps/chosen": -1.048828125, | |
| "logps/rejected": -2.979687452316284, | |
| "loss": 1.0622, | |
| "nll_loss": 1.0535156726837158, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05246581882238388, | |
| "rewards/margins": 0.096435546875, | |
| "rewards/rejected": -0.14897461235523224, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3096774193548387, | |
| "grad_norm": 4.227111777252394, | |
| "learning_rate": 6.46846227353151e-06, | |
| "log_odds_chosen": 2.953125, | |
| "log_odds_ratio": -0.14106445014476776, | |
| "logits/chosen": -2.9921875, | |
| "logits/rejected": 13.850000381469727, | |
| "logps/chosen": -1.001953125, | |
| "logps/rejected": -3.5546875, | |
| "loss": 1.0255, | |
| "nll_loss": 1.01953125, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05009765550494194, | |
| "rewards/margins": 0.12763671576976776, | |
| "rewards/rejected": -0.17763671278953552, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 4.690135110410384, | |
| "learning_rate": 6.337242505244779e-06, | |
| "log_odds_chosen": 3.825000047683716, | |
| "log_odds_ratio": -0.08366699516773224, | |
| "logits/chosen": -3.262890577316284, | |
| "logits/rejected": 14.893750190734863, | |
| "logps/chosen": -0.973828136920929, | |
| "logps/rejected": -4.339062690734863, | |
| "loss": 0.9886, | |
| "nll_loss": 0.9839843511581421, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04865722730755806, | |
| "rewards/margins": 0.16806641221046448, | |
| "rewards/rejected": -0.21699218451976776, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.33548387096774196, | |
| "grad_norm": 3.892806516901723, | |
| "learning_rate": 6.213697660012001e-06, | |
| "log_odds_chosen": 3.746875047683716, | |
| "log_odds_ratio": -0.08420410007238388, | |
| "logits/chosen": -2.832812547683716, | |
| "logits/rejected": 14.925000190734863, | |
| "logps/chosen": -0.987500011920929, | |
| "logps/rejected": -4.28125, | |
| "loss": 1.0076, | |
| "nll_loss": 1.00390625, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04938964918255806, | |
| "rewards/margins": 0.16474609076976776, | |
| "rewards/rejected": -0.21406249701976776, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.34838709677419355, | |
| "grad_norm": 4.635677028308263, | |
| "learning_rate": 6.097107608496925e-06, | |
| "log_odds_chosen": 3.6265625953674316, | |
| "log_odds_ratio": -0.10944823920726776, | |
| "logits/chosen": -2.586132764816284, | |
| "logits/rejected": 14.806249618530273, | |
| "logps/chosen": -0.9632812738418579, | |
| "logps/rejected": -4.150000095367432, | |
| "loss": 0.9866, | |
| "nll_loss": 0.9820312261581421, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04819335788488388, | |
| "rewards/margins": 0.15937499701976776, | |
| "rewards/rejected": -0.20742186903953552, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.36129032258064514, | |
| "grad_norm": 4.6072687211037975, | |
| "learning_rate": 5.986843400892497e-06, | |
| "log_odds_chosen": 3.8375000953674316, | |
| "log_odds_ratio": -0.08769531548023224, | |
| "logits/chosen": -2.3558592796325684, | |
| "logits/rejected": 15.043749809265137, | |
| "logps/chosen": -1.0109374523162842, | |
| "logps/rejected": -4.404687404632568, | |
| "loss": 1.0176, | |
| "nll_loss": 1.013281226158142, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05056152492761612, | |
| "rewards/margins": 0.169921875, | |
| "rewards/rejected": -0.22041015326976776, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3741935483870968, | |
| "grad_norm": 4.415703053607297, | |
| "learning_rate": 5.882352941176471e-06, | |
| "log_odds_chosen": 3.729687452316284, | |
| "log_odds_ratio": -0.08544921875, | |
| "logits/chosen": -2.450390577316284, | |
| "logits/rejected": 15.418749809265137, | |
| "logps/chosen": -0.974609375, | |
| "logps/rejected": -4.240624904632568, | |
| "loss": 0.9867, | |
| "nll_loss": 0.9839843511581421, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04873047024011612, | |
| "rewards/margins": 0.16318359971046448, | |
| "rewards/rejected": -0.21201172471046448, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 5.690639188843815, | |
| "learning_rate": 5.783149319662402e-06, | |
| "log_odds_chosen": 4.370312690734863, | |
| "log_odds_ratio": -0.07833252102136612, | |
| "logits/chosen": -2.629687547683716, | |
| "logits/rejected": 15.225000381469727, | |
| "logps/chosen": -1.0046875476837158, | |
| "logps/rejected": -4.921875, | |
| "loss": 1.0159, | |
| "nll_loss": 1.011328101158142, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05024414137005806, | |
| "rewards/margins": 0.19580078125, | |
| "rewards/rejected": -0.24599608778953552, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.012778448745233, | |
| "learning_rate": 5.688801239885744e-06, | |
| "log_odds_chosen": 4.082812309265137, | |
| "log_odds_ratio": -0.08937988430261612, | |
| "logits/chosen": -2.50390625, | |
| "logits/rejected": 14.6875, | |
| "logps/chosen": -0.979296863079071, | |
| "logps/rejected": -4.614062309265137, | |
| "loss": 0.997, | |
| "nll_loss": 0.993359386920929, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04897461086511612, | |
| "rewards/margins": 0.18159179389476776, | |
| "rewards/rejected": -0.23046875, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4129032258064516, | |
| "grad_norm": 4.657589224007925, | |
| "learning_rate": 5.598925109558544e-06, | |
| "log_odds_chosen": 4.184374809265137, | |
| "log_odds_ratio": -0.07525634765625, | |
| "logits/chosen": -2.4296875, | |
| "logits/rejected": 15.256250381469727, | |
| "logps/chosen": -0.9761718511581421, | |
| "logps/rejected": -4.701562404632568, | |
| "loss": 0.984, | |
| "nll_loss": 0.979687511920929, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.048828125, | |
| "rewards/margins": 0.18593749403953552, | |
| "rewards/rejected": -0.23476561903953552, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4258064516129032, | |
| "grad_norm": 5.589904480370191, | |
| "learning_rate": 5.513178464199713e-06, | |
| "log_odds_chosen": 4.478125095367432, | |
| "log_odds_ratio": -0.06633301079273224, | |
| "logits/chosen": -2.649609327316284, | |
| "logits/rejected": 15.387499809265137, | |
| "logps/chosen": -0.999218761920929, | |
| "logps/rejected": -5.017187595367432, | |
| "loss": 1.028, | |
| "nll_loss": 1.025781273841858, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.0499267578125, | |
| "rewards/margins": 0.20087890326976776, | |
| "rewards/rejected": -0.2509765625, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.43870967741935485, | |
| "grad_norm": 4.6064011071073026, | |
| "learning_rate": 5.4312544659356846e-06, | |
| "log_odds_chosen": 4.3046875, | |
| "log_odds_ratio": -0.08154296875, | |
| "logits/chosen": -2.225781202316284, | |
| "logits/rejected": 15.181249618530273, | |
| "logps/chosen": -1.0089843273162842, | |
| "logps/rejected": -4.871874809265137, | |
| "loss": 1.0052, | |
| "nll_loss": 1.002343773841858, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05046386644244194, | |
| "rewards/margins": 0.19287109375, | |
| "rewards/rejected": -0.24326172471046448, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.45161290322580644, | |
| "grad_norm": 4.5221044579303085, | |
| "learning_rate": 5.352877275721893e-06, | |
| "log_odds_chosen": 4.354687690734863, | |
| "log_odds_ratio": -0.08122558891773224, | |
| "logits/chosen": -2.8828125, | |
| "logits/rejected": 15.024999618530273, | |
| "logps/chosen": -0.9515625238418579, | |
| "logps/rejected": -4.828125, | |
| "loss": 0.9809, | |
| "nll_loss": 0.977734386920929, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04758300632238388, | |
| "rewards/margins": 0.19404296576976776, | |
| "rewards/rejected": -0.24150390923023224, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4645161290322581, | |
| "grad_norm": 5.034689811028297, | |
| "learning_rate": 5.277798139692596e-06, | |
| "log_odds_chosen": 4.385937690734863, | |
| "log_odds_ratio": -0.07797851413488388, | |
| "logits/chosen": -2.65234375, | |
| "logits/rejected": 15.131250381469727, | |
| "logps/chosen": -0.974609375, | |
| "logps/rejected": -4.896874904632568, | |
| "loss": 0.9937, | |
| "nll_loss": 0.991015613079071, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.04868163913488388, | |
| "rewards/margins": 0.19609375298023224, | |
| "rewards/rejected": -0.24501952528953552, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4774193548387097, | |
| "grad_norm": 14.430473695913276, | |
| "learning_rate": 5.205792062953535e-06, | |
| "log_odds_chosen": 4.551562309265137, | |
| "log_odds_ratio": -0.06669922173023224, | |
| "logits/chosen": -3.0826172828674316, | |
| "logits/rejected": 15.375, | |
| "logps/chosen": -1.000390648841858, | |
| "logps/rejected": -5.084374904632568, | |
| "loss": 1.0033, | |
| "nll_loss": 1.000390648841858, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04997558519244194, | |
| "rewards/margins": 0.20419922471046448, | |
| "rewards/rejected": -0.25439453125, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.49032258064516127, | |
| "grad_norm": 4.022201188727424, | |
| "learning_rate": 5.136654969375142e-06, | |
| "log_odds_chosen": 3.5140624046325684, | |
| "log_odds_ratio": -0.1080322265625, | |
| "logits/chosen": -2.8984375, | |
| "logits/rejected": 15.037500381469727, | |
| "logps/chosen": -0.962109386920929, | |
| "logps/rejected": -4.021874904632568, | |
| "loss": 0.9682, | |
| "nll_loss": 0.9632812738418579, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04819335788488388, | |
| "rewards/margins": 0.15278320014476776, | |
| "rewards/rejected": -0.20107421278953552, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5032258064516129, | |
| "grad_norm": 4.6129488372536995, | |
| "learning_rate": 5.070201265633938e-06, | |
| "log_odds_chosen": 3.4203124046325684, | |
| "log_odds_ratio": -0.0938720703125, | |
| "logits/chosen": -3.04296875, | |
| "logits/rejected": 15.25, | |
| "logps/chosen": -0.952343761920929, | |
| "logps/rejected": -3.9156250953674316, | |
| "loss": 0.9822, | |
| "nll_loss": 0.977734386920929, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04758300632238388, | |
| "rewards/margins": 0.14799804985523224, | |
| "rewards/rejected": -0.19550780951976776, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 4.444893820100578, | |
| "learning_rate": 5.006261743217589e-06, | |
| "log_odds_chosen": 3.379687547683716, | |
| "log_odds_ratio": -0.07974853366613388, | |
| "logits/chosen": -3.168750047683716, | |
| "logits/rejected": 14.993749618530273, | |
| "logps/chosen": -0.9671875238418579, | |
| "logps/rejected": -3.8812499046325684, | |
| "loss": 0.9826, | |
| "nll_loss": 0.9800781011581421, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04838867112994194, | |
| "rewards/margins": 0.14565429091453552, | |
| "rewards/rejected": -0.19394531846046448, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5290322580645161, | |
| "grad_norm": 4.906317709773465, | |
| "learning_rate": 4.944681764341488e-06, | |
| "log_odds_chosen": 3.46875, | |
| "log_odds_ratio": -0.067138671875, | |
| "logits/chosen": -2.827343702316284, | |
| "logits/rejected": 15.118749618530273, | |
| "logps/chosen": -0.966796875, | |
| "logps/rejected": -3.965625047683716, | |
| "loss": 0.9761, | |
| "nll_loss": 0.97265625, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04833984375, | |
| "rewards/margins": 0.15009765326976776, | |
| "rewards/rejected": -0.19843749701976776, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5419354838709678, | |
| "grad_norm": 4.844755914165153, | |
| "learning_rate": 4.885319687460315e-06, | |
| "log_odds_chosen": 3.5374999046325684, | |
| "log_odds_ratio": -0.08986816555261612, | |
| "logits/chosen": -2.745312452316284, | |
| "logits/rejected": 14.856249809265137, | |
| "logps/chosen": -0.981640636920929, | |
| "logps/rejected": -4.0703125, | |
| "loss": 0.9941, | |
| "nll_loss": 0.989453136920929, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.049072265625, | |
| "rewards/margins": 0.15439453721046448, | |
| "rewards/rejected": -0.20361328125, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5548387096774193, | |
| "grad_norm": 4.640648526664558, | |
| "learning_rate": 4.8280454958526765e-06, | |
| "log_odds_chosen": 3.7046875953674316, | |
| "log_odds_ratio": -0.07432861626148224, | |
| "logits/chosen": -1.938574194908142, | |
| "logits/rejected": 15.850000381469727, | |
| "logps/chosen": -0.9945312738418579, | |
| "logps/rejected": -4.234375, | |
| "loss": 0.9885, | |
| "nll_loss": 0.985156238079071, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04975586012005806, | |
| "rewards/margins": 0.162109375, | |
| "rewards/rejected": -0.21181640028953552, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.567741935483871, | |
| "grad_norm": 5.651687778086319, | |
| "learning_rate": 4.7727395990334775e-06, | |
| "log_odds_chosen": 3.496875047683716, | |
| "log_odds_ratio": -0.09119872748851776, | |
| "logits/chosen": -3.115307569503784, | |
| "logits/rejected": 15.087499618530273, | |
| "logps/chosen": -0.981640636920929, | |
| "logps/rejected": -4.03125, | |
| "loss": 1.0027, | |
| "nll_loss": 0.998046875, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04912109300494194, | |
| "rewards/margins": 0.15253905951976776, | |
| "rewards/rejected": -0.20185546576976776, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5806451612903226, | |
| "grad_norm": 4.282703734054553, | |
| "learning_rate": 4.719291781830088e-06, | |
| "log_odds_chosen": 3.354687452316284, | |
| "log_odds_ratio": -0.12017822265625, | |
| "logits/chosen": -2.7964844703674316, | |
| "logits/rejected": 14.881250381469727, | |
| "logps/chosen": -1.001562476158142, | |
| "logps/rejected": -3.9312500953674316, | |
| "loss": 1.0237, | |
| "nll_loss": 1.017578125, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.05007324367761612, | |
| "rewards/margins": 0.14653320610523224, | |
| "rewards/rejected": -0.19648437201976776, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5935483870967742, | |
| "grad_norm": 4.666237310414398, | |
| "learning_rate": 4.667600280093366e-06, | |
| "log_odds_chosen": 3.6953125, | |
| "log_odds_ratio": -0.09528808295726776, | |
| "logits/chosen": -2.721484422683716, | |
| "logits/rejected": 15.293749809265137, | |
| "logps/chosen": -0.9652343988418579, | |
| "logps/rejected": -4.1953125, | |
| "loss": 0.9829, | |
| "nll_loss": 0.977734386920929, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.04829101637005806, | |
| "rewards/margins": 0.16171875596046448, | |
| "rewards/rejected": -0.21005859971046448, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6064516129032258, | |
| "grad_norm": 5.24217311652083, | |
| "learning_rate": 4.6175709653961015e-06, | |
| "log_odds_chosen": 3.479687452316284, | |
| "log_odds_ratio": -0.116455078125, | |
| "logits/chosen": -3.159374952316284, | |
| "logits/rejected": 14.5625, | |
| "logps/chosen": -1.0109374523162842, | |
| "logps/rejected": -4.078125, | |
| "loss": 1.0153, | |
| "nll_loss": 1.009765625, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.05058593675494194, | |
| "rewards/margins": 0.15336914360523224, | |
| "rewards/rejected": -0.20390625298023224, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6193548387096774, | |
| "grad_norm": 4.85775451265865, | |
| "learning_rate": 4.569116623849539e-06, | |
| "log_odds_chosen": 3.8921875953674316, | |
| "log_odds_ratio": -0.07180175930261612, | |
| "logits/chosen": -2.889843702316284, | |
| "logits/rejected": 15.493749618530273, | |
| "logps/chosen": -0.922656238079071, | |
| "logps/rejected": -4.332812309265137, | |
| "loss": 0.9427, | |
| "nll_loss": 0.939453125, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04611816257238388, | |
| "rewards/margins": 0.17031249403953552, | |
| "rewards/rejected": -0.21640625596046448, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.632258064516129, | |
| "grad_norm": 5.135467312058317, | |
| "learning_rate": 4.522156316461347e-06, | |
| "log_odds_chosen": 3.7578125, | |
| "log_odds_ratio": -0.07160644233226776, | |
| "logits/chosen": -3.319140672683716, | |
| "logits/rejected": 15.512499809265137, | |
| "logps/chosen": -0.998828113079071, | |
| "logps/rejected": -4.303124904632568, | |
| "loss": 1.011, | |
| "nll_loss": 1.0078125, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04987793043255806, | |
| "rewards/margins": 0.1650390625, | |
| "rewards/rejected": -0.21494141221046448, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 4.47674714460599, | |
| "learning_rate": 4.476614810358452e-06, | |
| "log_odds_chosen": 3.5999999046325684, | |
| "log_odds_ratio": -0.09897460788488388, | |
| "logits/chosen": -2.8734374046325684, | |
| "logits/rejected": 15.068750381469727, | |
| "logps/chosen": -1.0261719226837158, | |
| "logps/rejected": -4.203125, | |
| "loss": 1.0312, | |
| "nll_loss": 1.0265624523162842, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05129394680261612, | |
| "rewards/margins": 0.15883788466453552, | |
| "rewards/rejected": -0.21005859971046448, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6580645161290323, | |
| "grad_norm": 4.714299310109519, | |
| "learning_rate": 4.432422071779363e-06, | |
| "log_odds_chosen": 3.8812499046325684, | |
| "log_odds_ratio": -0.07042236626148224, | |
| "logits/chosen": -3.11328125, | |
| "logits/rejected": 14.949999809265137, | |
| "logps/chosen": -0.936328113079071, | |
| "logps/rejected": -4.3359375, | |
| "loss": 0.9671, | |
| "nll_loss": 0.964062511920929, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04677734524011612, | |
| "rewards/margins": 0.16997070610523224, | |
| "rewards/rejected": -0.216796875, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6709677419354839, | |
| "grad_norm": 4.695855571935763, | |
| "learning_rate": 4.389512813061471e-06, | |
| "log_odds_chosen": 3.796875, | |
| "log_odds_ratio": -0.07448730617761612, | |
| "logits/chosen": -3.2523436546325684, | |
| "logits/rejected": 15.300000190734863, | |
| "logps/chosen": -0.931640625, | |
| "logps/rejected": -4.240624904632568, | |
| "loss": 0.9521, | |
| "nll_loss": 0.948437511920929, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04655761644244194, | |
| "rewards/margins": 0.16542968153953552, | |
| "rewards/rejected": -0.21171875298023224, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6838709677419355, | |
| "grad_norm": 4.8627800661273435, | |
| "learning_rate": 4.347826086956522e-06, | |
| "log_odds_chosen": 3.7015624046325684, | |
| "log_odds_ratio": -0.08720703423023224, | |
| "logits/chosen": -2.1195311546325684, | |
| "logits/rejected": 14.824999809265137, | |
| "logps/chosen": -0.930859386920929, | |
| "logps/rejected": -4.146874904632568, | |
| "loss": 0.9535, | |
| "nll_loss": 0.949999988079071, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.0465087890625, | |
| "rewards/margins": 0.16103515028953552, | |
| "rewards/rejected": -0.20751953125, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6967741935483871, | |
| "grad_norm": 4.178975744588359, | |
| "learning_rate": 4.3073049225394805e-06, | |
| "log_odds_chosen": 3.934375047683716, | |
| "log_odds_ratio": -0.08649902045726776, | |
| "logits/chosen": -2.1851563453674316, | |
| "logits/rejected": 15.850000381469727, | |
| "logps/chosen": -0.992968738079071, | |
| "logps/rejected": -4.431250095367432, | |
| "loss": 0.9628, | |
| "nll_loss": 0.958984375, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.049560546875, | |
| "rewards/margins": 0.17207030951976776, | |
| "rewards/rejected": -0.22187499701976776, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7096774193548387, | |
| "grad_norm": 5.086676612389064, | |
| "learning_rate": 4.2678959977632e-06, | |
| "log_odds_chosen": 3.9453125, | |
| "log_odds_ratio": -0.05517578125, | |
| "logits/chosen": -2.5117430686950684, | |
| "logits/rejected": 15.34375, | |
| "logps/chosen": -0.971484363079071, | |
| "logps/rejected": -4.439062595367432, | |
| "loss": 0.9747, | |
| "nll_loss": 0.97265625, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.048583984375, | |
| "rewards/margins": 0.17333984375, | |
| "rewards/rejected": -0.22197265923023224, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7225806451612903, | |
| "grad_norm": 4.675506919842173, | |
| "learning_rate": 4.229549344378136e-06, | |
| "log_odds_chosen": 4.035937309265137, | |
| "log_odds_ratio": -0.07060547173023224, | |
| "logits/chosen": -3.0234375, | |
| "logits/rejected": 15.137499809265137, | |
| "logps/chosen": -0.9703124761581421, | |
| "logps/rejected": -4.532812595367432, | |
| "loss": 0.9885, | |
| "nll_loss": 0.9839843511581421, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.0484619140625, | |
| "rewards/margins": 0.17822265625, | |
| "rewards/rejected": -0.22675780951976776, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7354838709677419, | |
| "grad_norm": 4.355789903905205, | |
| "learning_rate": 4.192218081503185e-06, | |
| "log_odds_chosen": 4.096875190734863, | |
| "log_odds_ratio": -0.0665283203125, | |
| "logits/chosen": -2.786328077316284, | |
| "logits/rejected": 15.324999809265137, | |
| "logps/chosen": -0.939453125, | |
| "logps/rejected": -4.553124904632568, | |
| "loss": 0.9593, | |
| "nll_loss": 0.9574218988418579, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04704589769244194, | |
| "rewards/margins": 0.18056640028953552, | |
| "rewards/rejected": -0.2275390625, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7483870967741936, | |
| "grad_norm": 4.395304577880807, | |
| "learning_rate": 4.155858174616914e-06, | |
| "log_odds_chosen": 3.8609375953674316, | |
| "log_odds_ratio": -0.06821288913488388, | |
| "logits/chosen": -3.245312452316284, | |
| "logits/rejected": 14.631250381469727, | |
| "logps/chosen": -0.912890613079071, | |
| "logps/rejected": -4.275000095367432, | |
| "loss": 0.9378, | |
| "nll_loss": 0.934765636920929, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.045654296875, | |
| "rewards/margins": 0.16787108778953552, | |
| "rewards/rejected": -0.21367187798023224, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7612903225806451, | |
| "grad_norm": 4.529044726069004, | |
| "learning_rate": 4.120428217151647e-06, | |
| "log_odds_chosen": 4.0546875, | |
| "log_odds_ratio": -0.06181640550494194, | |
| "logits/chosen": -3.321093797683716, | |
| "logits/rejected": 14.987500190734863, | |
| "logps/chosen": -0.928906261920929, | |
| "logps/rejected": -4.490624904632568, | |
| "loss": 0.9545, | |
| "nll_loss": 0.953125, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04638671875, | |
| "rewards/margins": 0.17802734673023224, | |
| "rewards/rejected": -0.22451171278953552, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 4.367330516996232, | |
| "learning_rate": 4.0858892322271844e-06, | |
| "log_odds_chosen": 4.139062404632568, | |
| "log_odds_ratio": -0.05810546875, | |
| "logits/chosen": -2.7339844703674316, | |
| "logits/rejected": 15.324999809265137, | |
| "logps/chosen": -0.9683593511581421, | |
| "logps/rejected": -4.634375095367432, | |
| "loss": 0.9833, | |
| "nll_loss": 0.98046875, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04836425930261612, | |
| "rewards/margins": 0.18339844048023224, | |
| "rewards/rejected": -0.23154297471046448, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7870967741935484, | |
| "grad_norm": 4.725621105355558, | |
| "learning_rate": 4.052204492365539e-06, | |
| "log_odds_chosen": 4.568749904632568, | |
| "log_odds_ratio": -0.05581054836511612, | |
| "logits/chosen": -3.828125, | |
| "logits/rejected": 15.168749809265137, | |
| "logps/chosen": -0.974609375, | |
| "logps/rejected": -5.065625190734863, | |
| "loss": 0.9903, | |
| "nll_loss": 0.98828125, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.0487060546875, | |
| "rewards/margins": 0.20458984375, | |
| "rewards/rejected": -0.25341796875, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 5.489269120536017, | |
| "learning_rate": 4.019339355290704e-06, | |
| "log_odds_chosen": 4.253125190734863, | |
| "log_odds_ratio": -0.08012695610523224, | |
| "logits/chosen": -3.4429688453674316, | |
| "logits/rejected": 15.5, | |
| "logps/chosen": -0.935546875, | |
| "logps/rejected": -4.696875095367432, | |
| "loss": 0.9602, | |
| "nll_loss": 0.9566406011581421, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.0467529296875, | |
| "rewards/margins": 0.18798828125, | |
| "rewards/rejected": -0.23466797173023224, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8129032258064516, | |
| "grad_norm": 4.806303543591708, | |
| "learning_rate": 3.9872611141445005e-06, | |
| "log_odds_chosen": 4.260937690734863, | |
| "log_odds_ratio": -0.05676879733800888, | |
| "logits/chosen": -3.21044921875, | |
| "logits/rejected": 15.40625, | |
| "logps/chosen": -0.919921875, | |
| "logps/rejected": -4.668749809265137, | |
| "loss": 0.9316, | |
| "nll_loss": 0.9292968511581421, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04604492336511612, | |
| "rewards/margins": 0.18740233778953552, | |
| "rewards/rejected": -0.2333984375, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8258064516129032, | |
| "grad_norm": 19.816287661530026, | |
| "learning_rate": 3.955938860646178e-06, | |
| "log_odds_chosen": 4.512499809265137, | |
| "log_odds_ratio": -0.04911499097943306, | |
| "logits/chosen": -2.807861328125, | |
| "logits/rejected": 15.481249809265137, | |
| "logps/chosen": -0.9945312738418579, | |
| "logps/rejected": -5.021874904632568, | |
| "loss": 0.9932, | |
| "nll_loss": 0.9917968511581421, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04973144456744194, | |
| "rewards/margins": 0.20107421278953552, | |
| "rewards/rejected": -0.2508789002895355, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8387096774193549, | |
| "grad_norm": 4.519655498618536, | |
| "learning_rate": 3.925343359894298e-06, | |
| "log_odds_chosen": 4.828125, | |
| "log_odds_ratio": -0.04646606370806694, | |
| "logits/chosen": -3.143749952316284, | |
| "logits/rejected": 15.987500190734863, | |
| "logps/chosen": -0.9808593988418579, | |
| "logps/rejected": -5.324999809265137, | |
| "loss": 0.9964, | |
| "nll_loss": 0.9937499761581421, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04909668117761612, | |
| "rewards/margins": 0.21708984673023224, | |
| "rewards/rejected": -0.26640623807907104, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8516129032258064, | |
| "grad_norm": 4.6066718377177915, | |
| "learning_rate": 3.895446935658099e-06, | |
| "log_odds_chosen": 4.415625095367432, | |
| "log_odds_ratio": -0.05909423902630806, | |
| "logits/chosen": -2.4632811546325684, | |
| "logits/rejected": 15.399999618530273, | |
| "logps/chosen": -0.987500011920929, | |
| "logps/rejected": -4.942187309265137, | |
| "loss": 0.9912, | |
| "nll_loss": 0.98828125, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04936523362994194, | |
| "rewards/margins": 0.19785156846046448, | |
| "rewards/rejected": -0.2470703125, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.864516129032258, | |
| "grad_norm": 4.613695358429598, | |
| "learning_rate": 3.866223365135295e-06, | |
| "log_odds_chosen": 4.578125, | |
| "log_odds_ratio": -0.05357055738568306, | |
| "logits/chosen": -3.272656202316284, | |
| "logits/rejected": 15.193750381469727, | |
| "logps/chosen": -0.9609375, | |
| "logps/rejected": -5.043749809265137, | |
| "loss": 0.9754, | |
| "nll_loss": 0.9730468988418579, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.0479736328125, | |
| "rewards/margins": 0.20439453423023224, | |
| "rewards/rejected": -0.25244140625, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.8774193548387097, | |
| "grad_norm": 5.422841177317399, | |
| "learning_rate": 3.837647782266681e-06, | |
| "log_odds_chosen": 4.890625, | |
| "log_odds_ratio": -0.0390625, | |
| "logits/chosen": -2.7265625, | |
| "logits/rejected": 15.375, | |
| "logps/chosen": -0.9867187738418579, | |
| "logps/rejected": -5.390625, | |
| "loss": 0.9915, | |
| "nll_loss": 0.989453136920929, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04931640625, | |
| "rewards/margins": 0.22021484375, | |
| "rewards/rejected": -0.2699218690395355, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8903225806451613, | |
| "grad_norm": 5.19065215935842, | |
| "learning_rate": 3.8096965887972973e-06, | |
| "log_odds_chosen": 4.612500190734863, | |
| "log_odds_ratio": -0.073883056640625, | |
| "logits/chosen": -2.76953125, | |
| "logits/rejected": 14.987500190734863, | |
| "logps/chosen": -0.9644531011581421, | |
| "logps/rejected": -5.084374904632568, | |
| "loss": 0.9617, | |
| "nll_loss": 0.957812488079071, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04824218899011612, | |
| "rewards/margins": 0.20585937798023224, | |
| "rewards/rejected": -0.25468748807907104, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 4.754599806255075, | |
| "learning_rate": 3.7823473723611694e-06, | |
| "log_odds_chosen": 4.498437404632568, | |
| "log_odds_ratio": -0.05502929538488388, | |
| "logits/chosen": -2.788281202316284, | |
| "logits/rejected": 15.15625, | |
| "logps/chosen": -0.923828125, | |
| "logps/rejected": -4.918749809265137, | |
| "loss": 0.9378, | |
| "nll_loss": 0.9359375238418579, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04621582105755806, | |
| "rewards/margins": 0.19970703125, | |
| "rewards/rejected": -0.24628905951976776, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9161290322580645, | |
| "grad_norm": 4.940556198099481, | |
| "learning_rate": 3.7555788309433988e-06, | |
| "log_odds_chosen": 4.724999904632568, | |
| "log_odds_ratio": -0.06513671576976776, | |
| "logits/chosen": -3.089062452316284, | |
| "logits/rejected": 15.15625, | |
| "logps/chosen": -0.9515625238418579, | |
| "logps/rejected": -5.196875095367432, | |
| "loss": 0.9742, | |
| "nll_loss": 0.971484363079071, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04763183742761612, | |
| "rewards/margins": 0.21220703423023224, | |
| "rewards/rejected": -0.2596679627895355, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9290322580645162, | |
| "grad_norm": 4.423887385405989, | |
| "learning_rate": 3.729370703141008e-06, | |
| "log_odds_chosen": 4.8828125, | |
| "log_odds_ratio": -0.04531250149011612, | |
| "logits/chosen": -2.8277344703674316, | |
| "logits/rejected": 15.324999809265137, | |
| "logps/chosen": -0.9242187738418579, | |
| "logps/rejected": -5.300000190734863, | |
| "loss": 0.9459, | |
| "nll_loss": 0.944531261920929, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.0462646484375, | |
| "rewards/margins": 0.21894530951976776, | |
| "rewards/rejected": -0.26494139432907104, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9419354838709677, | |
| "grad_norm": 5.674218130480263, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "log_odds_chosen": 4.359375, | |
| "log_odds_ratio": -0.05672607570886612, | |
| "logits/chosen": -3.0914063453674316, | |
| "logits/rejected": 15.118749618530273, | |
| "logps/chosen": -0.95703125, | |
| "logps/rejected": -4.837500095367432, | |
| "loss": 0.9725, | |
| "nll_loss": 0.9691406488418579, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04782714694738388, | |
| "rewards/margins": 0.19394531846046448, | |
| "rewards/rejected": -0.24160155653953552, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.9548387096774194, | |
| "grad_norm": 4.565990551466217, | |
| "learning_rate": 3.6785594638885484e-06, | |
| "log_odds_chosen": 4.143750190734863, | |
| "log_odds_ratio": -0.06418456882238388, | |
| "logits/chosen": -3.5328125953674316, | |
| "logits/rejected": 14.65625, | |
| "logps/chosen": -0.9203125238418579, | |
| "logps/rejected": -4.557812690734863, | |
| "loss": 0.9248, | |
| "nll_loss": 0.921875, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04599609225988388, | |
| "rewards/margins": 0.18203124403953552, | |
| "rewards/rejected": -0.22812500596046448, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 4.4761450056044705, | |
| "learning_rate": 3.6539204762093485e-06, | |
| "log_odds_chosen": 4.503125190734863, | |
| "log_odds_ratio": -0.04804076999425888, | |
| "logits/chosen": -3.145312547683716, | |
| "logits/rejected": 15.393750190734863, | |
| "logps/chosen": -0.958984375, | |
| "logps/rejected": -4.959374904632568, | |
| "loss": 0.9787, | |
| "nll_loss": 0.9769531488418579, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04794921725988388, | |
| "rewards/margins": 0.20000000298023224, | |
| "rewards/rejected": -0.248046875, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9806451612903225, | |
| "grad_norm": 4.612842620007858, | |
| "learning_rate": 3.6297700432031386e-06, | |
| "log_odds_chosen": 4.284375190734863, | |
| "log_odds_ratio": -0.07587890326976776, | |
| "logits/chosen": -2.9046874046325684, | |
| "logits/rejected": 15.106249809265137, | |
| "logps/chosen": -0.950390636920929, | |
| "logps/rejected": -4.759375095367432, | |
| "loss": 0.969, | |
| "nll_loss": 0.96484375, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04750976711511612, | |
| "rewards/margins": 0.19003906846046448, | |
| "rewards/rejected": -0.23789063096046448, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.9935483870967742, | |
| "grad_norm": 5.433862964765063, | |
| "learning_rate": 3.606092229873094e-06, | |
| "log_odds_chosen": 4.482812404632568, | |
| "log_odds_ratio": -0.07089843600988388, | |
| "logits/chosen": -3.5140624046325684, | |
| "logits/rejected": 14.943750381469727, | |
| "logps/chosen": -0.9488281011581421, | |
| "logps/rejected": -4.940625190734863, | |
| "loss": 0.9684, | |
| "nll_loss": 0.9652343988418579, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04746093600988388, | |
| "rewards/margins": 0.19970703125, | |
| "rewards/rejected": -0.24746093153953552, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 775, | |
| "total_flos": 0.0, | |
| "train_loss": 1.0403899949596773, | |
| "train_runtime": 1570.4069, | |
| "train_samples_per_second": 7.896, | |
| "train_steps_per_second": 0.494 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 775, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |