| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.99968, | |
| "eval_steps": 300, | |
| "global_step": 781, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 38.75, | |
| "learning_rate": 6.329113924050633e-09, | |
| "logits/generated": 0.1862616091966629, | |
| "logits/real": -0.24196800589561462, | |
| "logps/generated": -538.626953125, | |
| "logps/real": -411.1020202636719, | |
| "loss": 0.5, | |
| "rewards/accuracies": 0.0, | |
| "rewards/generated": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/real": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 32.0, | |
| "learning_rate": 3.1645569620253166e-08, | |
| "logits/generated": 0.34831082820892334, | |
| "logits/real": -0.12466613948345184, | |
| "logps/generated": -534.7514038085938, | |
| "logps/real": -401.6441345214844, | |
| "loss": 0.4933, | |
| "rewards/accuracies": 0.390625, | |
| "rewards/generated": 0.028874225914478302, | |
| "rewards/margins": -0.008529577404260635, | |
| "rewards/real": 0.020344652235507965, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 29.125, | |
| "learning_rate": 6.329113924050633e-08, | |
| "logits/generated": 0.37132272124290466, | |
| "logits/real": -0.12087633460760117, | |
| "logps/generated": -555.05859375, | |
| "logps/real": -418.5079040527344, | |
| "loss": 0.4941, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/generated": 0.00217894883826375, | |
| "rewards/margins": -0.003474087920039892, | |
| "rewards/real": -0.0012951357057318091, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 31.375, | |
| "learning_rate": 9.49367088607595e-08, | |
| "logits/generated": 0.3862614035606384, | |
| "logits/real": -0.1023840680718422, | |
| "logps/generated": -564.0845947265625, | |
| "logps/real": -435.2381286621094, | |
| "loss": 0.4815, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/generated": -0.025657888501882553, | |
| "rewards/margins": 0.04708629846572876, | |
| "rewards/real": 0.021428411826491356, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 30.75, | |
| "learning_rate": 1.2658227848101266e-07, | |
| "logits/generated": 0.3665524125099182, | |
| "logits/real": -0.06457304954528809, | |
| "logps/generated": -558.3909301757812, | |
| "logps/real": -439.05072021484375, | |
| "loss": 0.5034, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/generated": 0.007058300077915192, | |
| "rewards/margins": -0.009866954758763313, | |
| "rewards/real": -0.0028086546808481216, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 31.5, | |
| "learning_rate": 1.5822784810126582e-07, | |
| "logits/generated": 0.39779749512672424, | |
| "logits/real": -0.1216607466340065, | |
| "logps/generated": -565.197509765625, | |
| "logps/real": -439.4720153808594, | |
| "loss": 0.4893, | |
| "rewards/accuracies": 0.5, | |
| "rewards/generated": 0.0025315589737147093, | |
| "rewards/margins": -0.010492642410099506, | |
| "rewards/real": -0.007961079478263855, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 32.5, | |
| "learning_rate": 1.89873417721519e-07, | |
| "logits/generated": 0.3631184995174408, | |
| "logits/real": -0.15178892016410828, | |
| "logps/generated": -563.345947265625, | |
| "logps/real": -444.08758544921875, | |
| "loss": 0.4822, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/generated": -0.02776198461651802, | |
| "rewards/margins": 0.05192292481660843, | |
| "rewards/real": 0.024160942062735558, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 32.5, | |
| "learning_rate": 2.2151898734177212e-07, | |
| "logits/generated": 0.353170782327652, | |
| "logits/real": -0.08487684279680252, | |
| "logps/generated": -567.288330078125, | |
| "logps/real": -426.352783203125, | |
| "loss": 0.478, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/generated": -0.05789921432733536, | |
| "rewards/margins": 0.08254577219486237, | |
| "rewards/real": 0.02464655414223671, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 28.0, | |
| "learning_rate": 2.5316455696202533e-07, | |
| "logits/generated": 0.3951062560081482, | |
| "logits/real": -0.10047361999750137, | |
| "logps/generated": -584.459228515625, | |
| "logps/real": -481.53759765625, | |
| "loss": 0.4788, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/generated": -0.01975923776626587, | |
| "rewards/margins": 0.056668005883693695, | |
| "rewards/real": 0.036908768117427826, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 29.375, | |
| "learning_rate": 2.848101265822785e-07, | |
| "logits/generated": 0.38055142760276794, | |
| "logits/real": -0.10685946047306061, | |
| "logps/generated": -590.1273193359375, | |
| "logps/real": -475.4049377441406, | |
| "loss": 0.4803, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/generated": -0.034241292625665665, | |
| "rewards/margins": 0.02389874868094921, | |
| "rewards/real": -0.010342543944716454, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 26.875, | |
| "learning_rate": 3.1645569620253163e-07, | |
| "logits/generated": 0.362557053565979, | |
| "logits/real": -0.13479682803153992, | |
| "logps/generated": -559.6375122070312, | |
| "logps/real": -434.8016662597656, | |
| "loss": 0.4702, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/generated": -0.0708869993686676, | |
| "rewards/margins": 0.08817923069000244, | |
| "rewards/real": 0.01729224994778633, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 33.25, | |
| "learning_rate": 3.481012658227848e-07, | |
| "logits/generated": 0.3691774010658264, | |
| "logits/real": -0.11431871354579926, | |
| "logps/generated": -591.8275146484375, | |
| "logps/real": -450.1565856933594, | |
| "loss": 0.4806, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/generated": -0.07000661641359329, | |
| "rewards/margins": 0.05859777331352234, | |
| "rewards/real": -0.011408851481974125, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 32.0, | |
| "learning_rate": 3.79746835443038e-07, | |
| "logits/generated": 0.3952616751194, | |
| "logits/real": -0.08973407000303268, | |
| "logps/generated": -569.9074096679688, | |
| "logps/real": -456.4561462402344, | |
| "loss": 0.4661, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/generated": -0.06687521934509277, | |
| "rewards/margins": 0.08400523662567139, | |
| "rewards/real": 0.017130006104707718, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 27.25, | |
| "learning_rate": 4.1139240506329117e-07, | |
| "logits/generated": 0.36575227975845337, | |
| "logits/real": -0.1133301854133606, | |
| "logps/generated": -542.1326293945312, | |
| "logps/real": -441.4345703125, | |
| "loss": 0.4689, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/generated": -0.07646136730909348, | |
| "rewards/margins": 0.08980376273393631, | |
| "rewards/real": 0.013342398218810558, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 29.875, | |
| "learning_rate": 4.4303797468354424e-07, | |
| "logits/generated": 0.3912655711174011, | |
| "logits/real": -0.1571967601776123, | |
| "logps/generated": -570.942626953125, | |
| "logps/real": -486.6839904785156, | |
| "loss": 0.4531, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/generated": -0.13936015963554382, | |
| "rewards/margins": 0.1767238825559616, | |
| "rewards/real": 0.03736373037099838, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 26.5, | |
| "learning_rate": 4.746835443037974e-07, | |
| "logits/generated": 0.3816757798194885, | |
| "logits/real": -0.09128452837467194, | |
| "logps/generated": -571.598876953125, | |
| "logps/real": -419.603515625, | |
| "loss": 0.4616, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/generated": -0.09458805620670319, | |
| "rewards/margins": 0.08172789961099625, | |
| "rewards/real": -0.012860171496868134, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 28.375, | |
| "learning_rate": 4.992877492877492e-07, | |
| "logits/generated": 0.34975099563598633, | |
| "logits/real": -0.14952535927295685, | |
| "logps/generated": -547.2381591796875, | |
| "logps/real": -442.2696838378906, | |
| "loss": 0.4441, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/generated": -0.13002917170524597, | |
| "rewards/margins": 0.15946698188781738, | |
| "rewards/real": 0.029437804594635963, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 28.75, | |
| "learning_rate": 4.957264957264958e-07, | |
| "logits/generated": 0.35858625173568726, | |
| "logits/real": -0.18670549988746643, | |
| "logps/generated": -535.11669921875, | |
| "logps/real": -456.5271911621094, | |
| "loss": 0.4408, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/generated": -0.18137022852897644, | |
| "rewards/margins": 0.22222211956977844, | |
| "rewards/real": 0.040851861238479614, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 29.5, | |
| "learning_rate": 4.921652421652421e-07, | |
| "logits/generated": 0.30655255913734436, | |
| "logits/real": -0.13807684183120728, | |
| "logps/generated": -579.0311279296875, | |
| "logps/real": -435.41485595703125, | |
| "loss": 0.4332, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/generated": -0.17220628261566162, | |
| "rewards/margins": 0.18943142890930176, | |
| "rewards/real": 0.01722514070570469, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 28.25, | |
| "learning_rate": 4.886039886039886e-07, | |
| "logits/generated": 0.3390645980834961, | |
| "logits/real": -0.159017413854599, | |
| "logps/generated": -562.1036376953125, | |
| "logps/real": -472.2007751464844, | |
| "loss": 0.4335, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/generated": -0.19639980792999268, | |
| "rewards/margins": 0.21172475814819336, | |
| "rewards/real": 0.015324937179684639, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 27.375, | |
| "learning_rate": 4.850427350427351e-07, | |
| "logits/generated": 0.3581392168998718, | |
| "logits/real": -0.22140832245349884, | |
| "logps/generated": -518.3876953125, | |
| "logps/real": -505.3263244628906, | |
| "loss": 0.4226, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/generated": -0.22456979751586914, | |
| "rewards/margins": 0.27873098850250244, | |
| "rewards/real": 0.0541611909866333, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 29.0, | |
| "learning_rate": 4.814814814814814e-07, | |
| "logits/generated": 0.3360639214515686, | |
| "logits/real": -0.16623374819755554, | |
| "logps/generated": -597.8626708984375, | |
| "logps/real": -469.96435546875, | |
| "loss": 0.4089, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/generated": -0.2624703645706177, | |
| "rewards/margins": 0.3305851221084595, | |
| "rewards/real": 0.06811471283435822, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 27.125, | |
| "learning_rate": 4.779202279202279e-07, | |
| "logits/generated": 0.3120338022708893, | |
| "logits/real": -0.1398223340511322, | |
| "logps/generated": -559.6121826171875, | |
| "logps/real": -435.9784240722656, | |
| "loss": 0.4151, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/generated": -0.2531023323535919, | |
| "rewards/margins": 0.2851903438568115, | |
| "rewards/real": 0.03208799287676811, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 27.0, | |
| "learning_rate": 4.743589743589743e-07, | |
| "logits/generated": 0.35134896636009216, | |
| "logits/real": -0.16236919164657593, | |
| "logps/generated": -540.6229248046875, | |
| "logps/real": -481.10479736328125, | |
| "loss": 0.3949, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/generated": -0.3117538094520569, | |
| "rewards/margins": 0.37438851594924927, | |
| "rewards/real": 0.06263472884893417, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 28.0, | |
| "learning_rate": 4.707977207977208e-07, | |
| "logits/generated": 0.31295281648635864, | |
| "logits/real": -0.15100577473640442, | |
| "logps/generated": -533.11376953125, | |
| "logps/real": -425.769775390625, | |
| "loss": 0.4001, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/generated": -0.3296114504337311, | |
| "rewards/margins": 0.37464508414268494, | |
| "rewards/real": 0.04503365978598595, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 28.375, | |
| "learning_rate": 4.672364672364672e-07, | |
| "logits/generated": 0.3123488128185272, | |
| "logits/real": -0.1613120287656784, | |
| "logps/generated": -556.4259643554688, | |
| "logps/real": -451.27593994140625, | |
| "loss": 0.3915, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/generated": -0.32520076632499695, | |
| "rewards/margins": 0.3783460259437561, | |
| "rewards/real": 0.05314524844288826, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 28.875, | |
| "learning_rate": 4.6367521367521367e-07, | |
| "logits/generated": 0.3258894085884094, | |
| "logits/real": -0.1628173589706421, | |
| "logps/generated": -558.4219360351562, | |
| "logps/real": -435.81243896484375, | |
| "loss": 0.3934, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/generated": -0.36564749479293823, | |
| "rewards/margins": 0.4104110300540924, | |
| "rewards/real": 0.04476350545883179, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 26.75, | |
| "learning_rate": 4.601139601139601e-07, | |
| "logits/generated": 0.2820456326007843, | |
| "logits/real": -0.2306266725063324, | |
| "logps/generated": -578.9539794921875, | |
| "logps/real": -447.5259704589844, | |
| "loss": 0.3839, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/generated": -0.3908230662345886, | |
| "rewards/margins": 0.4163753390312195, | |
| "rewards/real": 0.02555229887366295, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 28.25, | |
| "learning_rate": 4.5655270655270654e-07, | |
| "logits/generated": 0.3293055295944214, | |
| "logits/real": -0.24442501366138458, | |
| "logps/generated": -557.1941528320312, | |
| "logps/real": -533.494873046875, | |
| "loss": 0.3766, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/generated": -0.37160658836364746, | |
| "rewards/margins": 0.4162415862083435, | |
| "rewards/real": 0.044634997844696045, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 28.0, | |
| "learning_rate": 4.5299145299145297e-07, | |
| "logits/generated": 0.28380945324897766, | |
| "logits/real": -0.16644850373268127, | |
| "logps/generated": -532.15625, | |
| "logps/real": -427.69287109375, | |
| "loss": 0.3824, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/generated": -0.42885756492614746, | |
| "rewards/margins": 0.4583609104156494, | |
| "rewards/real": 0.029503341764211655, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 26.375, | |
| "learning_rate": 4.494301994301994e-07, | |
| "logits/generated": 0.26398712396621704, | |
| "logits/real": -0.1775752604007721, | |
| "logps/generated": -544.0301513671875, | |
| "logps/real": -442.7901916503906, | |
| "loss": 0.3883, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/generated": -0.4080891013145447, | |
| "rewards/margins": 0.4677355885505676, | |
| "rewards/real": 0.05964647978544235, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 28.125, | |
| "learning_rate": 4.4586894586894584e-07, | |
| "logits/generated": 0.3235207796096802, | |
| "logits/real": -0.19184669852256775, | |
| "logps/generated": -565.9249267578125, | |
| "logps/real": -485.18109130859375, | |
| "loss": 0.3713, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/generated": -0.43480604887008667, | |
| "rewards/margins": 0.4802734851837158, | |
| "rewards/real": 0.04546744376420975, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 27.875, | |
| "learning_rate": 4.423076923076923e-07, | |
| "logits/generated": 0.3015090823173523, | |
| "logits/real": -0.2252935916185379, | |
| "logps/generated": -566.9320068359375, | |
| "logps/real": -444.4369201660156, | |
| "loss": 0.3552, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/generated": -0.45722684264183044, | |
| "rewards/margins": 0.554154634475708, | |
| "rewards/real": 0.09692778438329697, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 26.5, | |
| "learning_rate": 4.3874643874643876e-07, | |
| "logits/generated": 0.31183922290802, | |
| "logits/real": -0.1933097243309021, | |
| "logps/generated": -558.2478637695312, | |
| "logps/real": -460.18267822265625, | |
| "loss": 0.3597, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/generated": -0.47245973348617554, | |
| "rewards/margins": 0.5111454129219055, | |
| "rewards/real": 0.03868564963340759, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 27.375, | |
| "learning_rate": 4.3518518518518514e-07, | |
| "logits/generated": 0.250281423330307, | |
| "logits/real": -0.1833321750164032, | |
| "logps/generated": -536.4636840820312, | |
| "logps/real": -416.641357421875, | |
| "loss": 0.364, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/generated": -0.454394668340683, | |
| "rewards/margins": 0.5187661647796631, | |
| "rewards/real": 0.06437145173549652, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 26.625, | |
| "learning_rate": 4.3162393162393163e-07, | |
| "logits/generated": 0.29415473341941833, | |
| "logits/real": -0.24404272437095642, | |
| "logps/generated": -538.131103515625, | |
| "logps/real": -477.87603759765625, | |
| "loss": 0.348, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.5133287310600281, | |
| "rewards/margins": 0.5939925909042358, | |
| "rewards/real": 0.08066388964653015, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 30.125, | |
| "learning_rate": 4.2806267806267807e-07, | |
| "logits/generated": 0.27652794122695923, | |
| "logits/real": -0.2590489685535431, | |
| "logps/generated": -561.6041259765625, | |
| "logps/real": -444.3932189941406, | |
| "loss": 0.3536, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/generated": -0.47919169068336487, | |
| "rewards/margins": 0.5749484896659851, | |
| "rewards/real": 0.09575681388378143, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 26.75, | |
| "learning_rate": 4.245014245014245e-07, | |
| "logits/generated": 0.3380230963230133, | |
| "logits/real": -0.11477069556713104, | |
| "logps/generated": -599.4484252929688, | |
| "logps/real": -421.1564025878906, | |
| "loss": 0.3357, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.5778478384017944, | |
| "rewards/margins": 0.6686479449272156, | |
| "rewards/real": 0.09080011397600174, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 27.0, | |
| "learning_rate": 4.2094017094017093e-07, | |
| "logits/generated": 0.3196483850479126, | |
| "logits/real": -0.1744014322757721, | |
| "logps/generated": -586.2642822265625, | |
| "logps/real": -426.1298828125, | |
| "loss": 0.3573, | |
| "rewards/accuracies": 0.875, | |
| "rewards/generated": -0.48544687032699585, | |
| "rewards/margins": 0.557063102722168, | |
| "rewards/real": 0.07161618769168854, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 27.125, | |
| "learning_rate": 4.173789173789173e-07, | |
| "logits/generated": 0.25956088304519653, | |
| "logits/real": -0.2154865711927414, | |
| "logps/generated": -528.0460205078125, | |
| "logps/real": -452.2561950683594, | |
| "loss": 0.3429, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/generated": -0.5490601658821106, | |
| "rewards/margins": 0.6324241757392883, | |
| "rewards/real": 0.08336395770311356, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 26.5, | |
| "learning_rate": 4.138176638176638e-07, | |
| "logits/generated": 0.3287113904953003, | |
| "logits/real": -0.20184390246868134, | |
| "logps/generated": -580.03857421875, | |
| "logps/real": -471.832763671875, | |
| "loss": 0.3437, | |
| "rewards/accuracies": 0.875, | |
| "rewards/generated": -0.5396500825881958, | |
| "rewards/margins": 0.575441837310791, | |
| "rewards/real": 0.03579169511795044, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 25.5, | |
| "learning_rate": 4.1025641025641024e-07, | |
| "logits/generated": 0.3303323984146118, | |
| "logits/real": -0.18564388155937195, | |
| "logps/generated": -576.2296752929688, | |
| "logps/real": -470.49554443359375, | |
| "loss": 0.3435, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/generated": -0.5153689384460449, | |
| "rewards/margins": 0.5805807113647461, | |
| "rewards/real": 0.06521172821521759, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 28.0, | |
| "learning_rate": 4.0669515669515667e-07, | |
| "logits/generated": 0.25469496846199036, | |
| "logits/real": -0.28663891553878784, | |
| "logps/generated": -558.517822265625, | |
| "logps/real": -467.5023498535156, | |
| "loss": 0.3327, | |
| "rewards/accuracies": 0.875, | |
| "rewards/generated": -0.5688196420669556, | |
| "rewards/margins": 0.6551179885864258, | |
| "rewards/real": 0.08629832416772842, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 23.75, | |
| "learning_rate": 4.031339031339031e-07, | |
| "logits/generated": 0.2868548333644867, | |
| "logits/real": -0.19922226667404175, | |
| "logps/generated": -549.5493774414062, | |
| "logps/real": -457.3771057128906, | |
| "loss": 0.3296, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/generated": -0.6514773964881897, | |
| "rewards/margins": 0.7027413845062256, | |
| "rewards/real": 0.05126393958926201, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 25.375, | |
| "learning_rate": 3.995726495726496e-07, | |
| "logits/generated": 0.3001692295074463, | |
| "logits/real": -0.18475447595119476, | |
| "logps/generated": -541.4611206054688, | |
| "logps/real": -444.6468200683594, | |
| "loss": 0.338, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/generated": -0.5863200426101685, | |
| "rewards/margins": 0.6246292591094971, | |
| "rewards/real": 0.03830911964178085, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 24.625, | |
| "learning_rate": 3.9601139601139597e-07, | |
| "logits/generated": 0.2538192868232727, | |
| "logits/real": -0.2353261411190033, | |
| "logps/generated": -573.8902587890625, | |
| "logps/real": -450.2127380371094, | |
| "loss": 0.3309, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/generated": -0.5875332951545715, | |
| "rewards/margins": 0.6803909540176392, | |
| "rewards/real": 0.09285767376422882, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 26.125, | |
| "learning_rate": 3.924501424501424e-07, | |
| "logits/generated": 0.2551348805427551, | |
| "logits/real": -0.2044079303741455, | |
| "logps/generated": -546.6871337890625, | |
| "logps/real": -421.7835998535156, | |
| "loss": 0.3181, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/generated": -0.5918381214141846, | |
| "rewards/margins": 0.6859561204910278, | |
| "rewards/real": 0.09411804378032684, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 25.375, | |
| "learning_rate": 3.888888888888889e-07, | |
| "logits/generated": 0.3060089349746704, | |
| "logits/real": -0.21332141757011414, | |
| "logps/generated": -597.86376953125, | |
| "logps/real": -443.6407165527344, | |
| "loss": 0.32, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/generated": -0.6655870676040649, | |
| "rewards/margins": 0.729436993598938, | |
| "rewards/real": 0.06384985148906708, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 26.125, | |
| "learning_rate": 3.853276353276353e-07, | |
| "logits/generated": 0.30169203877449036, | |
| "logits/real": -0.1919897347688675, | |
| "logps/generated": -541.0394287109375, | |
| "logps/real": -423.79833984375, | |
| "loss": 0.3223, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -0.5991157293319702, | |
| "rewards/margins": 0.68681800365448, | |
| "rewards/real": 0.08770229667425156, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 25.375, | |
| "learning_rate": 3.8176638176638176e-07, | |
| "logits/generated": 0.2983294129371643, | |
| "logits/real": -0.23591478168964386, | |
| "logps/generated": -559.3455200195312, | |
| "logps/real": -503.95751953125, | |
| "loss": 0.3408, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/generated": -0.6077674031257629, | |
| "rewards/margins": 0.6536551713943481, | |
| "rewards/real": 0.04588776454329491, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 27.125, | |
| "learning_rate": 3.782051282051282e-07, | |
| "logits/generated": 0.26214712858200073, | |
| "logits/real": -0.2555038034915924, | |
| "logps/generated": -555.8446044921875, | |
| "logps/real": -438.34423828125, | |
| "loss": 0.3161, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.6663080453872681, | |
| "rewards/margins": 0.738936185836792, | |
| "rewards/real": 0.07262810319662094, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 24.75, | |
| "learning_rate": 3.7464387464387463e-07, | |
| "logits/generated": 0.30381038784980774, | |
| "logits/real": -0.2225998193025589, | |
| "logps/generated": -538.4495849609375, | |
| "logps/real": -484.36993408203125, | |
| "loss": 0.3231, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/generated": -0.6721928715705872, | |
| "rewards/margins": 0.7017067670822144, | |
| "rewards/real": 0.02951394021511078, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 25.5, | |
| "learning_rate": 3.7108262108262107e-07, | |
| "logits/generated": 0.28403228521347046, | |
| "logits/real": -0.24081599712371826, | |
| "logps/generated": -552.0731811523438, | |
| "logps/real": -438.84356689453125, | |
| "loss": 0.3093, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/generated": -0.6770501732826233, | |
| "rewards/margins": 0.7596859335899353, | |
| "rewards/real": 0.08263583481311798, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 26.0, | |
| "learning_rate": 3.6752136752136755e-07, | |
| "logits/generated": 0.2751469016075134, | |
| "logits/real": -0.2667590081691742, | |
| "logps/generated": -574.9193115234375, | |
| "logps/real": -458.193359375, | |
| "loss": 0.3005, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -0.6883004307746887, | |
| "rewards/margins": 0.8165790438652039, | |
| "rewards/real": 0.1282786875963211, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 28.125, | |
| "learning_rate": 3.6396011396011393e-07, | |
| "logits/generated": 0.3092981278896332, | |
| "logits/real": -0.24193081259727478, | |
| "logps/generated": -537.2657470703125, | |
| "logps/real": -463.10784912109375, | |
| "loss": 0.3048, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -0.6616233587265015, | |
| "rewards/margins": 0.76463383436203, | |
| "rewards/real": 0.10301046073436737, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 25.0, | |
| "learning_rate": 3.6039886039886037e-07, | |
| "logits/generated": 0.29445740580558777, | |
| "logits/real": -0.21788445115089417, | |
| "logps/generated": -562.7876586914062, | |
| "logps/real": -432.4056701660156, | |
| "loss": 0.2912, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.7921088337898254, | |
| "rewards/margins": 0.866392970085144, | |
| "rewards/real": 0.07428419589996338, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 28.125, | |
| "learning_rate": 3.5683760683760686e-07, | |
| "logits/generated": 0.29365968704223633, | |
| "logits/real": -0.16456809639930725, | |
| "logps/generated": -568.531005859375, | |
| "logps/real": -452.4728088378906, | |
| "loss": 0.2857, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.7912765145301819, | |
| "rewards/margins": 0.9027554392814636, | |
| "rewards/real": 0.11147890239953995, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 29.875, | |
| "learning_rate": 3.5327635327635324e-07, | |
| "logits/generated": 0.2611847519874573, | |
| "logits/real": -0.23247161507606506, | |
| "logps/generated": -557.5487060546875, | |
| "logps/real": -437.1328125, | |
| "loss": 0.301, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.7435336112976074, | |
| "rewards/margins": 0.8042283058166504, | |
| "rewards/real": 0.0606946162879467, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 26.5, | |
| "learning_rate": 3.497150997150997e-07, | |
| "logits/generated": 0.2905731797218323, | |
| "logits/real": -0.29508811235427856, | |
| "logps/generated": -568.8480834960938, | |
| "logps/real": -493.10662841796875, | |
| "loss": 0.298, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.8001793622970581, | |
| "rewards/margins": 0.8642541766166687, | |
| "rewards/real": 0.06407480686903, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 25.125, | |
| "learning_rate": 3.461538461538461e-07, | |
| "logits/generated": 0.25090640783309937, | |
| "logits/real": -0.23970182240009308, | |
| "logps/generated": -571.603759765625, | |
| "logps/real": -419.73162841796875, | |
| "loss": 0.289, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.8356046676635742, | |
| "rewards/margins": 0.9479316473007202, | |
| "rewards/real": 0.11232707649469376, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 23.0, | |
| "learning_rate": 3.425925925925926e-07, | |
| "logits/generated": 0.24801941215991974, | |
| "logits/real": -0.2514159381389618, | |
| "logps/generated": -578.6694946289062, | |
| "logps/real": -446.20550537109375, | |
| "loss": 0.2932, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/generated": -0.7759395837783813, | |
| "rewards/margins": 0.8407967686653137, | |
| "rewards/real": 0.06485719978809357, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_logits/generated": 0.25535792112350464, | |
| "eval_logits/real": -0.2529524266719818, | |
| "eval_logps/generated": -568.39892578125, | |
| "eval_logps/real": -468.0555725097656, | |
| "eval_loss": 0.29618778824806213, | |
| "eval_rewards/accuracies": 0.921999990940094, | |
| "eval_rewards/generated": -0.7854698300361633, | |
| "eval_rewards/margins": 0.857230007648468, | |
| "eval_rewards/real": 0.07176008075475693, | |
| "eval_runtime": 251.8666, | |
| "eval_samples_per_second": 7.941, | |
| "eval_steps_per_second": 0.993, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 27.375, | |
| "learning_rate": 3.3903133903133903e-07, | |
| "logits/generated": 0.24424830079078674, | |
| "logits/real": -0.2476576864719391, | |
| "logps/generated": -556.5860595703125, | |
| "logps/real": -447.5859375, | |
| "loss": 0.2816, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.7970044612884521, | |
| "rewards/margins": 0.8752428889274597, | |
| "rewards/real": 0.0782383605837822, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 24.875, | |
| "learning_rate": 3.354700854700854e-07, | |
| "logits/generated": 0.25422555208206177, | |
| "logits/real": -0.27572309970855713, | |
| "logps/generated": -581.782958984375, | |
| "logps/real": -469.9566955566406, | |
| "loss": 0.2844, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -0.8562766313552856, | |
| "rewards/margins": 0.9663234949111938, | |
| "rewards/real": 0.11004678159952164, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 23.875, | |
| "learning_rate": 3.319088319088319e-07, | |
| "logits/generated": 0.2173062115907669, | |
| "logits/real": -0.2551296651363373, | |
| "logps/generated": -550.9298706054688, | |
| "logps/real": -478.9058532714844, | |
| "loss": 0.3036, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.7514694333076477, | |
| "rewards/margins": 0.7880412340164185, | |
| "rewards/real": 0.03657188266515732, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 25.625, | |
| "learning_rate": 3.2834757834757833e-07, | |
| "logits/generated": 0.27889084815979004, | |
| "logits/real": -0.19512683153152466, | |
| "logps/generated": -569.74560546875, | |
| "logps/real": -422.5086364746094, | |
| "loss": 0.2891, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.8094115257263184, | |
| "rewards/margins": 0.8925756216049194, | |
| "rewards/real": 0.08316410332918167, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 25.125, | |
| "learning_rate": 3.2478632478632476e-07, | |
| "logits/generated": 0.2879219949245453, | |
| "logits/real": -0.27906739711761475, | |
| "logps/generated": -583.55908203125, | |
| "logps/real": -479.68988037109375, | |
| "loss": 0.2798, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.7989848852157593, | |
| "rewards/margins": 0.9030762910842896, | |
| "rewards/real": 0.10409130901098251, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 23.125, | |
| "learning_rate": 3.212250712250712e-07, | |
| "logits/generated": 0.2541460394859314, | |
| "logits/real": -0.21789531409740448, | |
| "logps/generated": -570.9312744140625, | |
| "logps/real": -437.07476806640625, | |
| "loss": 0.2908, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.8303514719009399, | |
| "rewards/margins": 0.9341219663619995, | |
| "rewards/real": 0.10377053916454315, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 25.875, | |
| "learning_rate": 3.176638176638177e-07, | |
| "logits/generated": 0.2649766802787781, | |
| "logits/real": -0.23893487453460693, | |
| "logps/generated": -601.1578979492188, | |
| "logps/real": -482.6011657714844, | |
| "loss": 0.2859, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -0.8597179651260376, | |
| "rewards/margins": 0.926683247089386, | |
| "rewards/real": 0.06696534156799316, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 26.75, | |
| "learning_rate": 3.1410256410256407e-07, | |
| "logits/generated": 0.2579233646392822, | |
| "logits/real": -0.2822721600532532, | |
| "logps/generated": -575.4134521484375, | |
| "logps/real": -461.46099853515625, | |
| "loss": 0.2751, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -0.8965667486190796, | |
| "rewards/margins": 0.9989431500434875, | |
| "rewards/real": 0.10237620025873184, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 24.125, | |
| "learning_rate": 3.1054131054131055e-07, | |
| "logits/generated": 0.24282464385032654, | |
| "logits/real": -0.19455142319202423, | |
| "logps/generated": -573.3221435546875, | |
| "logps/real": -401.08319091796875, | |
| "loss": 0.2932, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.7664304971694946, | |
| "rewards/margins": 0.8475470542907715, | |
| "rewards/real": 0.08111655712127686, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 21.875, | |
| "learning_rate": 3.06980056980057e-07, | |
| "logits/generated": 0.2672547698020935, | |
| "logits/real": -0.2481343299150467, | |
| "logps/generated": -546.6334228515625, | |
| "logps/real": -449.42694091796875, | |
| "loss": 0.2879, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.8173602223396301, | |
| "rewards/margins": 0.8795957565307617, | |
| "rewards/real": 0.062235575169324875, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 23.25, | |
| "learning_rate": 3.0341880341880337e-07, | |
| "logits/generated": 0.2704509198665619, | |
| "logits/real": -0.26537832617759705, | |
| "logps/generated": -572.9697265625, | |
| "logps/real": -483.6675720214844, | |
| "loss": 0.2776, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -0.8736934661865234, | |
| "rewards/margins": 0.9758409261703491, | |
| "rewards/real": 0.10214747488498688, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 24.375, | |
| "learning_rate": 2.9985754985754986e-07, | |
| "logits/generated": 0.23115357756614685, | |
| "logits/real": -0.21211636066436768, | |
| "logps/generated": -565.6680908203125, | |
| "logps/real": -419.73004150390625, | |
| "loss": 0.2911, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -0.8500432968139648, | |
| "rewards/margins": 0.9019571542739868, | |
| "rewards/real": 0.05191392824053764, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 24.0, | |
| "learning_rate": 2.962962962962963e-07, | |
| "logits/generated": 0.24056752026081085, | |
| "logits/real": -0.26113012433052063, | |
| "logps/generated": -570.0296020507812, | |
| "logps/real": -483.2586975097656, | |
| "loss": 0.2824, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.8894678354263306, | |
| "rewards/margins": 0.985159695148468, | |
| "rewards/real": 0.09569184482097626, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 24.75, | |
| "learning_rate": 2.927350427350427e-07, | |
| "logits/generated": 0.23835477232933044, | |
| "logits/real": -0.2653237283229828, | |
| "logps/generated": -572.784912109375, | |
| "logps/real": -469.38995361328125, | |
| "loss": 0.2879, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.9159420132637024, | |
| "rewards/margins": 0.9773737192153931, | |
| "rewards/real": 0.061431728303432465, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 21.25, | |
| "learning_rate": 2.8917378917378916e-07, | |
| "logits/generated": 0.23210255801677704, | |
| "logits/real": -0.18087713420391083, | |
| "logps/generated": -559.0404663085938, | |
| "logps/real": -409.7191467285156, | |
| "loss": 0.2834, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.861811637878418, | |
| "rewards/margins": 0.9838217496871948, | |
| "rewards/real": 0.12201017141342163, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 23.375, | |
| "learning_rate": 2.8561253561253565e-07, | |
| "logits/generated": 0.21506235003471375, | |
| "logits/real": -0.26088595390319824, | |
| "logps/generated": -571.0586547851562, | |
| "logps/real": -450.5771484375, | |
| "loss": 0.2905, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.8820127248764038, | |
| "rewards/margins": 0.9079700708389282, | |
| "rewards/real": 0.025957267731428146, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 23.0, | |
| "learning_rate": 2.8205128205128203e-07, | |
| "logits/generated": 0.22029800713062286, | |
| "logits/real": -0.25589895248413086, | |
| "logps/generated": -549.1640625, | |
| "logps/real": -407.2776794433594, | |
| "loss": 0.2792, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/generated": -0.8181031942367554, | |
| "rewards/margins": 0.9209718704223633, | |
| "rewards/real": 0.1028686910867691, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 23.125, | |
| "learning_rate": 2.7849002849002846e-07, | |
| "logits/generated": 0.2413676679134369, | |
| "logits/real": -0.2968316078186035, | |
| "logps/generated": -567.1182861328125, | |
| "logps/real": -498.888671875, | |
| "loss": 0.2883, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.8102291822433472, | |
| "rewards/margins": 0.8724681735038757, | |
| "rewards/real": 0.062238991260528564, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 24.375, | |
| "learning_rate": 2.749287749287749e-07, | |
| "logits/generated": 0.25530534982681274, | |
| "logits/real": -0.27374374866485596, | |
| "logps/generated": -560.980712890625, | |
| "logps/real": -457.9765625, | |
| "loss": 0.2639, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -0.9177068471908569, | |
| "rewards/margins": 1.0232750177383423, | |
| "rewards/real": 0.10556807368993759, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 23.5, | |
| "learning_rate": 2.7136752136752133e-07, | |
| "logits/generated": 0.2365385740995407, | |
| "logits/real": -0.2824167311191559, | |
| "logps/generated": -593.1279296875, | |
| "logps/real": -420.54608154296875, | |
| "loss": 0.2769, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.9472834467887878, | |
| "rewards/margins": 1.0492109060287476, | |
| "rewards/real": 0.1019275039434433, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 23.5, | |
| "learning_rate": 2.678062678062678e-07, | |
| "logits/generated": 0.2417086362838745, | |
| "logits/real": -0.21366417407989502, | |
| "logps/generated": -576.4993896484375, | |
| "logps/real": -450.2381896972656, | |
| "loss": 0.2767, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.8732357025146484, | |
| "rewards/margins": 0.9638237953186035, | |
| "rewards/real": 0.0905880555510521, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 24.375, | |
| "learning_rate": 2.642450142450142e-07, | |
| "logits/generated": 0.2748910188674927, | |
| "logits/real": -0.2791479229927063, | |
| "logps/generated": -584.0994262695312, | |
| "logps/real": -474.29632568359375, | |
| "loss": 0.2624, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/generated": -0.9621374011039734, | |
| "rewards/margins": 1.0574219226837158, | |
| "rewards/real": 0.09528439491987228, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 22.0, | |
| "learning_rate": 2.606837606837607e-07, | |
| "logits/generated": 0.24150626361370087, | |
| "logits/real": -0.2777560353279114, | |
| "logps/generated": -580.0728759765625, | |
| "logps/real": -422.99273681640625, | |
| "loss": 0.27, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/generated": -0.893054187297821, | |
| "rewards/margins": 1.0188217163085938, | |
| "rewards/real": 0.12576748430728912, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 23.125, | |
| "learning_rate": 2.571225071225071e-07, | |
| "logits/generated": 0.2542496621608734, | |
| "logits/real": -0.19691842794418335, | |
| "logps/generated": -562.2733154296875, | |
| "logps/real": -425.752685546875, | |
| "loss": 0.2695, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.9875919222831726, | |
| "rewards/margins": 1.0910958051681519, | |
| "rewards/real": 0.10350396484136581, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 23.125, | |
| "learning_rate": 2.5356125356125355e-07, | |
| "logits/generated": 0.20040163397789001, | |
| "logits/real": -0.2452303171157837, | |
| "logps/generated": -573.46142578125, | |
| "logps/real": -419.46795654296875, | |
| "loss": 0.264, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.9656850099563599, | |
| "rewards/margins": 1.063493013381958, | |
| "rewards/real": 0.0978081002831459, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 24.875, | |
| "learning_rate": 2.5e-07, | |
| "logits/generated": 0.2523021697998047, | |
| "logits/real": -0.2764134407043457, | |
| "logps/generated": -564.8800048828125, | |
| "logps/real": -534.353759765625, | |
| "loss": 0.2699, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/generated": -0.856059730052948, | |
| "rewards/margins": 0.9356383085250854, | |
| "rewards/real": 0.07957860082387924, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 25.75, | |
| "learning_rate": 2.464387464387464e-07, | |
| "logits/generated": 0.24086351692676544, | |
| "logits/real": -0.3009343147277832, | |
| "logps/generated": -557.8231201171875, | |
| "logps/real": -470.4691467285156, | |
| "loss": 0.2697, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.9576956629753113, | |
| "rewards/margins": 1.0484102964401245, | |
| "rewards/real": 0.09071467071771622, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 23.375, | |
| "learning_rate": 2.4287749287749286e-07, | |
| "logits/generated": 0.25982385873794556, | |
| "logits/real": -0.27899274230003357, | |
| "logps/generated": -618.6469116210938, | |
| "logps/real": -502.1651916503906, | |
| "loss": 0.269, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.9170626401901245, | |
| "rewards/margins": 1.0099645853042603, | |
| "rewards/real": 0.09290206432342529, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 21.5, | |
| "learning_rate": 2.393162393162393e-07, | |
| "logits/generated": 0.21476399898529053, | |
| "logits/real": -0.2801568806171417, | |
| "logps/generated": -586.9801025390625, | |
| "logps/real": -428.6065979003906, | |
| "loss": 0.2744, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.8958868980407715, | |
| "rewards/margins": 0.9335384368896484, | |
| "rewards/real": 0.03765156865119934, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 21.125, | |
| "learning_rate": 2.3575498575498575e-07, | |
| "logits/generated": 0.22820791602134705, | |
| "logits/real": -0.2373519241809845, | |
| "logps/generated": -591.01416015625, | |
| "logps/real": -455.0992736816406, | |
| "loss": 0.261, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.9810103178024292, | |
| "rewards/margins": 1.0630879402160645, | |
| "rewards/real": 0.08207770437002182, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 23.625, | |
| "learning_rate": 2.3219373219373216e-07, | |
| "logits/generated": 0.2243388593196869, | |
| "logits/real": -0.3200889527797699, | |
| "logps/generated": -562.8119506835938, | |
| "logps/real": -481.96697998046875, | |
| "loss": 0.2646, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -0.9419108629226685, | |
| "rewards/margins": 1.0310115814208984, | |
| "rewards/real": 0.08910069614648819, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 22.625, | |
| "learning_rate": 2.2863247863247862e-07, | |
| "logits/generated": 0.24365122616291046, | |
| "logits/real": -0.2204679548740387, | |
| "logps/generated": -557.0007934570312, | |
| "logps/real": -446.1201171875, | |
| "loss": 0.2557, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/generated": -0.9837862849235535, | |
| "rewards/margins": 1.0443240404129028, | |
| "rewards/real": 0.060537658631801605, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 24.5, | |
| "learning_rate": 2.2507122507122505e-07, | |
| "logits/generated": 0.21367163956165314, | |
| "logits/real": -0.26951321959495544, | |
| "logps/generated": -555.830078125, | |
| "logps/real": -465.67864990234375, | |
| "loss": 0.2615, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -1.0059828758239746, | |
| "rewards/margins": 1.0972858667373657, | |
| "rewards/real": 0.09130293130874634, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 23.125, | |
| "learning_rate": 2.215099715099715e-07, | |
| "logits/generated": 0.22326333820819855, | |
| "logits/real": -0.266071081161499, | |
| "logps/generated": -554.1658325195312, | |
| "logps/real": -408.9630432128906, | |
| "loss": 0.2685, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -0.8925862312316895, | |
| "rewards/margins": 0.9420120120048523, | |
| "rewards/real": 0.04942578822374344, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 23.25, | |
| "learning_rate": 2.1794871794871795e-07, | |
| "logits/generated": 0.24654094874858856, | |
| "logits/real": -0.2880334258079529, | |
| "logps/generated": -571.1951904296875, | |
| "logps/real": -466.7237854003906, | |
| "loss": 0.2653, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/generated": -0.9611663818359375, | |
| "rewards/margins": 1.0481185913085938, | |
| "rewards/real": 0.08695221692323685, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 21.375, | |
| "learning_rate": 2.1438746438746438e-07, | |
| "logits/generated": 0.2374725341796875, | |
| "logits/real": -0.2841905653476715, | |
| "logps/generated": -597.453125, | |
| "logps/real": -467.695068359375, | |
| "loss": 0.2606, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -1.0608772039413452, | |
| "rewards/margins": 1.1110285520553589, | |
| "rewards/real": 0.05015145614743233, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 22.625, | |
| "learning_rate": 2.1082621082621082e-07, | |
| "logits/generated": 0.20874783396720886, | |
| "logits/real": -0.24746723473072052, | |
| "logps/generated": -558.3968505859375, | |
| "logps/real": -418.06085205078125, | |
| "loss": 0.2779, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/generated": -0.9692691564559937, | |
| "rewards/margins": 1.016911268234253, | |
| "rewards/real": 0.04764218255877495, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 21.5, | |
| "learning_rate": 2.0726495726495728e-07, | |
| "logits/generated": 0.23431365191936493, | |
| "logits/real": -0.3022805154323578, | |
| "logps/generated": -560.6641845703125, | |
| "logps/real": -508.0726013183594, | |
| "loss": 0.2748, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -0.9711412191390991, | |
| "rewards/margins": 0.9574298858642578, | |
| "rewards/real": -0.013711368665099144, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 21.875, | |
| "learning_rate": 2.0370370370370369e-07, | |
| "logits/generated": 0.217031329870224, | |
| "logits/real": -0.2614264488220215, | |
| "logps/generated": -559.6924438476562, | |
| "logps/real": -446.6640625, | |
| "loss": 0.2687, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.9257688522338867, | |
| "rewards/margins": 0.9959582090377808, | |
| "rewards/real": 0.07018940150737762, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 22.625, | |
| "learning_rate": 2.0014245014245012e-07, | |
| "logits/generated": 0.24876682460308075, | |
| "logits/real": -0.28565463423728943, | |
| "logps/generated": -548.2100830078125, | |
| "logps/real": -472.73223876953125, | |
| "loss": 0.2661, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.9158692359924316, | |
| "rewards/margins": 1.0267049074172974, | |
| "rewards/real": 0.11083561182022095, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.9658119658119656e-07, | |
| "logits/generated": 0.21912440657615662, | |
| "logits/real": -0.27414244413375854, | |
| "logps/generated": -540.6039428710938, | |
| "logps/real": -458.55975341796875, | |
| "loss": 0.2633, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.9050576090812683, | |
| "rewards/margins": 1.0112102031707764, | |
| "rewards/real": 0.10615254938602448, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 23.375, | |
| "learning_rate": 1.9301994301994302e-07, | |
| "logits/generated": 0.22622957825660706, | |
| "logits/real": -0.25896841287612915, | |
| "logps/generated": -567.3508911132812, | |
| "logps/real": -443.44915771484375, | |
| "loss": 0.263, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.9187369346618652, | |
| "rewards/margins": 1.0117876529693604, | |
| "rewards/real": 0.09305067360401154, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 24.125, | |
| "learning_rate": 1.8945868945868945e-07, | |
| "logits/generated": 0.22547177970409393, | |
| "logits/real": -0.30368226766586304, | |
| "logps/generated": -568.8248291015625, | |
| "logps/real": -470.2293395996094, | |
| "loss": 0.2686, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -1.009271502494812, | |
| "rewards/margins": 1.0391043424606323, | |
| "rewards/real": 0.029832839965820312, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 23.625, | |
| "learning_rate": 1.8589743589743588e-07, | |
| "logits/generated": 0.228875070810318, | |
| "logits/real": -0.27519601583480835, | |
| "logps/generated": -548.3033447265625, | |
| "logps/real": -456.94049072265625, | |
| "loss": 0.2579, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/generated": -0.9734387397766113, | |
| "rewards/margins": 1.06186842918396, | |
| "rewards/real": 0.08842964470386505, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 22.25, | |
| "learning_rate": 1.8233618233618234e-07, | |
| "logits/generated": 0.24410124123096466, | |
| "logits/real": -0.24580614268779755, | |
| "logps/generated": -526.7844848632812, | |
| "logps/real": -461.39599609375, | |
| "loss": 0.2555, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.9268838763237, | |
| "rewards/margins": 1.0250850915908813, | |
| "rewards/real": 0.09820131957530975, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.7877492877492878e-07, | |
| "logits/generated": 0.24622035026550293, | |
| "logits/real": -0.350544810295105, | |
| "logps/generated": -583.2838134765625, | |
| "logps/real": -484.83758544921875, | |
| "loss": 0.2696, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -1.0305290222167969, | |
| "rewards/margins": 1.0726959705352783, | |
| "rewards/real": 0.04216689243912697, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 20.625, | |
| "learning_rate": 1.752136752136752e-07, | |
| "logits/generated": 0.23964126408100128, | |
| "logits/real": -0.21912190318107605, | |
| "logps/generated": -574.9818115234375, | |
| "logps/real": -413.7064514160156, | |
| "loss": 0.2533, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -1.0674055814743042, | |
| "rewards/margins": 1.1347934007644653, | |
| "rewards/real": 0.06738778948783875, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 21.375, | |
| "learning_rate": 1.7165242165242165e-07, | |
| "logits/generated": 0.2101718634366989, | |
| "logits/real": -0.2580256760120392, | |
| "logps/generated": -591.0747680664062, | |
| "logps/real": -402.9712219238281, | |
| "loss": 0.2586, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/generated": -0.9934548139572144, | |
| "rewards/margins": 1.0756797790527344, | |
| "rewards/real": 0.08222486823797226, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.6809116809116808e-07, | |
| "logits/generated": 0.1858968287706375, | |
| "logits/real": -0.28893381357192993, | |
| "logps/generated": -551.6287841796875, | |
| "logps/real": -424.68243408203125, | |
| "loss": 0.2614, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/generated": -0.9353107213973999, | |
| "rewards/margins": 1.0308815240859985, | |
| "rewards/real": 0.09557078033685684, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 22.0, | |
| "learning_rate": 1.6452991452991452e-07, | |
| "logits/generated": 0.23100645840168, | |
| "logits/real": -0.2850993275642395, | |
| "logps/generated": -597.5736083984375, | |
| "logps/real": -406.21990966796875, | |
| "loss": 0.2458, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -1.1093555688858032, | |
| "rewards/margins": 1.2105884552001953, | |
| "rewards/real": 0.1012328714132309, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 22.375, | |
| "learning_rate": 1.6096866096866095e-07, | |
| "logits/generated": 0.22843961417675018, | |
| "logits/real": -0.2411879003047943, | |
| "logps/generated": -565.7466430664062, | |
| "logps/real": -450.0619201660156, | |
| "loss": 0.2582, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -1.010749340057373, | |
| "rewards/margins": 1.1109167337417603, | |
| "rewards/real": 0.10016737133264542, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 22.25, | |
| "learning_rate": 1.574074074074074e-07, | |
| "logits/generated": 0.24150070548057556, | |
| "logits/real": -0.21642279624938965, | |
| "logps/generated": -569.0734252929688, | |
| "logps/real": -448.4789123535156, | |
| "loss": 0.2589, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/generated": -1.0423076152801514, | |
| "rewards/margins": 1.118851900100708, | |
| "rewards/real": 0.07654404640197754, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 23.625, | |
| "learning_rate": 1.5384615384615385e-07, | |
| "logits/generated": 0.18384118378162384, | |
| "logits/real": -0.2612574100494385, | |
| "logps/generated": -577.3256225585938, | |
| "logps/real": -445.0957946777344, | |
| "loss": 0.2617, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -1.027761697769165, | |
| "rewards/margins": 1.0843464136123657, | |
| "rewards/real": 0.056584786623716354, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 21.625, | |
| "learning_rate": 1.5028490028490028e-07, | |
| "logits/generated": 0.21284691989421844, | |
| "logits/real": -0.27072662115097046, | |
| "logps/generated": -576.1101684570312, | |
| "logps/real": -405.94073486328125, | |
| "loss": 0.2469, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -0.9797765612602234, | |
| "rewards/margins": 1.1207729578018188, | |
| "rewards/real": 0.14099633693695068, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 18.625, | |
| "learning_rate": 1.4672364672364671e-07, | |
| "logits/generated": 0.2502863109111786, | |
| "logits/real": -0.2541927099227905, | |
| "logps/generated": -583.0015869140625, | |
| "logps/real": -480.0306701660156, | |
| "loss": 0.2581, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.9324277639389038, | |
| "rewards/margins": 1.014402151107788, | |
| "rewards/real": 0.08197424560785294, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 22.25, | |
| "learning_rate": 1.4316239316239315e-07, | |
| "logits/generated": 0.17801856994628906, | |
| "logits/real": -0.29988187551498413, | |
| "logps/generated": -559.722412109375, | |
| "logps/real": -396.5543518066406, | |
| "loss": 0.2549, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -1.0840636491775513, | |
| "rewards/margins": 1.1837726831436157, | |
| "rewards/real": 0.09970910847187042, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 23.375, | |
| "learning_rate": 1.3960113960113958e-07, | |
| "logits/generated": 0.19342169165611267, | |
| "logits/real": -0.30865079164505005, | |
| "logps/generated": -531.6026611328125, | |
| "logps/real": -461.54766845703125, | |
| "loss": 0.2586, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.9994919896125793, | |
| "rewards/margins": 1.0923388004302979, | |
| "rewards/real": 0.09284672886133194, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 24.5, | |
| "learning_rate": 1.3603988603988604e-07, | |
| "logits/generated": 0.20833972096443176, | |
| "logits/real": -0.302274614572525, | |
| "logps/generated": -581.9873046875, | |
| "logps/real": -479.486572265625, | |
| "loss": 0.2588, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/generated": -1.043798565864563, | |
| "rewards/margins": 1.113098382949829, | |
| "rewards/real": 0.06929990649223328, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 22.25, | |
| "learning_rate": 1.3247863247863248e-07, | |
| "logits/generated": 0.18650811910629272, | |
| "logits/real": -0.2915360927581787, | |
| "logps/generated": -559.4674682617188, | |
| "logps/real": -441.85272216796875, | |
| "loss": 0.2401, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -1.024379849433899, | |
| "rewards/margins": 1.136856198310852, | |
| "rewards/real": 0.11247648298740387, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 20.875, | |
| "learning_rate": 1.289173789173789e-07, | |
| "logits/generated": 0.2449146956205368, | |
| "logits/real": -0.25022393465042114, | |
| "logps/generated": -576.7535400390625, | |
| "logps/real": -425.8819885253906, | |
| "loss": 0.2689, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/generated": -0.9874047040939331, | |
| "rewards/margins": 1.0526171922683716, | |
| "rewards/real": 0.06521251052618027, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_logits/generated": 0.22532238066196442, | |
| "eval_logits/real": -0.2819841802120209, | |
| "eval_logps/generated": -570.8114013671875, | |
| "eval_logps/real": -468.16961669921875, | |
| "eval_loss": 0.2591499090194702, | |
| "eval_rewards/accuracies": 0.9459999799728394, | |
| "eval_rewards/generated": -1.026713252067566, | |
| "eval_rewards/margins": 1.0870678424835205, | |
| "eval_rewards/real": 0.06035461276769638, | |
| "eval_runtime": 312.5704, | |
| "eval_samples_per_second": 6.399, | |
| "eval_steps_per_second": 0.8, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 20.125, | |
| "learning_rate": 1.2535612535612535e-07, | |
| "logits/generated": 0.26240265369415283, | |
| "logits/real": -0.27648916840553284, | |
| "logps/generated": -587.2186279296875, | |
| "logps/real": -469.5171813964844, | |
| "loss": 0.2429, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -1.1691229343414307, | |
| "rewards/margins": 1.2391382455825806, | |
| "rewards/real": 0.07001528143882751, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 20.625, | |
| "learning_rate": 1.2179487179487178e-07, | |
| "logits/generated": 0.22506709396839142, | |
| "logits/real": -0.24815583229064941, | |
| "logps/generated": -573.8429565429688, | |
| "logps/real": -422.6187438964844, | |
| "loss": 0.2596, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -1.049294352531433, | |
| "rewards/margins": 1.1830122470855713, | |
| "rewards/real": 0.13371792435646057, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 20.75, | |
| "learning_rate": 1.1823361823361823e-07, | |
| "logits/generated": 0.2133275717496872, | |
| "logits/real": -0.35235756635665894, | |
| "logps/generated": -569.0250854492188, | |
| "logps/real": -522.1539306640625, | |
| "loss": 0.2442, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/generated": -1.0183016061782837, | |
| "rewards/margins": 1.1286038160324097, | |
| "rewards/real": 0.11030229181051254, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 22.25, | |
| "learning_rate": 1.1467236467236467e-07, | |
| "logits/generated": 0.21725332736968994, | |
| "logits/real": -0.31801334023475647, | |
| "logps/generated": -525.4303588867188, | |
| "logps/real": -458.35955810546875, | |
| "loss": 0.2522, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -1.0983850955963135, | |
| "rewards/margins": 1.1680142879486084, | |
| "rewards/real": 0.06962917000055313, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 22.625, | |
| "learning_rate": 1.111111111111111e-07, | |
| "logits/generated": 0.19476717710494995, | |
| "logits/real": -0.2784198224544525, | |
| "logps/generated": -539.5494384765625, | |
| "logps/real": -444.2561950683594, | |
| "loss": 0.2603, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -1.0229387283325195, | |
| "rewards/margins": 1.1077196598052979, | |
| "rewards/real": 0.08478085696697235, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 21.875, | |
| "learning_rate": 1.0754985754985754e-07, | |
| "logits/generated": 0.1993536651134491, | |
| "logits/real": -0.27307894825935364, | |
| "logps/generated": -559.6204833984375, | |
| "logps/real": -425.73779296875, | |
| "loss": 0.2509, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -1.0144376754760742, | |
| "rewards/margins": 1.1288973093032837, | |
| "rewards/real": 0.11445963382720947, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 21.375, | |
| "learning_rate": 1.0398860398860399e-07, | |
| "logits/generated": 0.21056529879570007, | |
| "logits/real": -0.24936561286449432, | |
| "logps/generated": -586.7345581054688, | |
| "logps/real": -403.81884765625, | |
| "loss": 0.247, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -1.08762526512146, | |
| "rewards/margins": 1.1714038848876953, | |
| "rewards/real": 0.08377879858016968, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 20.125, | |
| "learning_rate": 1.0042735042735042e-07, | |
| "logits/generated": 0.22550058364868164, | |
| "logits/real": -0.20315103232860565, | |
| "logps/generated": -574.2731323242188, | |
| "logps/real": -442.43280029296875, | |
| "loss": 0.2638, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/generated": -1.087454080581665, | |
| "rewards/margins": 1.1403470039367676, | |
| "rewards/real": 0.05289285629987717, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 21.375, | |
| "learning_rate": 9.686609686609686e-08, | |
| "logits/generated": 0.2513348460197449, | |
| "logits/real": -0.24446964263916016, | |
| "logps/generated": -574.2137451171875, | |
| "logps/real": -461.9295959472656, | |
| "loss": 0.2621, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -0.9981436729431152, | |
| "rewards/margins": 1.0288463830947876, | |
| "rewards/real": 0.03070276975631714, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 21.25, | |
| "learning_rate": 9.33048433048433e-08, | |
| "logits/generated": 0.26496636867523193, | |
| "logits/real": -0.2916685938835144, | |
| "logps/generated": -551.1077880859375, | |
| "logps/real": -462.594482421875, | |
| "loss": 0.2516, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -1.0148528814315796, | |
| "rewards/margins": 1.0718117952346802, | |
| "rewards/real": 0.05695882439613342, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 23.125, | |
| "learning_rate": 8.974358974358974e-08, | |
| "logits/generated": 0.2397506982088089, | |
| "logits/real": -0.2423442304134369, | |
| "logps/generated": -585.54248046875, | |
| "logps/real": -456.861328125, | |
| "loss": 0.2585, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -1.0295922756195068, | |
| "rewards/margins": 1.0833699703216553, | |
| "rewards/real": 0.05377759784460068, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 20.625, | |
| "learning_rate": 8.618233618233619e-08, | |
| "logits/generated": 0.22315570712089539, | |
| "logits/real": -0.32908838987350464, | |
| "logps/generated": -574.5935668945312, | |
| "logps/real": -443.726806640625, | |
| "loss": 0.2626, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -1.041495680809021, | |
| "rewards/margins": 1.098388433456421, | |
| "rewards/real": 0.056892819702625275, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 22.375, | |
| "learning_rate": 8.262108262108261e-08, | |
| "logits/generated": 0.22458979487419128, | |
| "logits/real": -0.26257410645484924, | |
| "logps/generated": -587.2811279296875, | |
| "logps/real": -409.606201171875, | |
| "loss": 0.2487, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -1.0680924654006958, | |
| "rewards/margins": 1.1297204494476318, | |
| "rewards/real": 0.06162794679403305, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 22.0, | |
| "learning_rate": 7.905982905982906e-08, | |
| "logits/generated": 0.2559250593185425, | |
| "logits/real": -0.26228904724121094, | |
| "logps/generated": -611.9027099609375, | |
| "logps/real": -435.93634033203125, | |
| "loss": 0.2567, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/generated": -1.0482969284057617, | |
| "rewards/margins": 1.1413335800170898, | |
| "rewards/real": 0.09303668141365051, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 22.0, | |
| "learning_rate": 7.549857549857549e-08, | |
| "logits/generated": 0.21574994921684265, | |
| "logits/real": -0.3184639811515808, | |
| "logps/generated": -589.6176147460938, | |
| "logps/real": -459.68212890625, | |
| "loss": 0.2653, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -1.039894700050354, | |
| "rewards/margins": 1.0526447296142578, | |
| "rewards/real": 0.012749890796840191, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 19.75, | |
| "learning_rate": 7.193732193732194e-08, | |
| "logits/generated": 0.2510204613208771, | |
| "logits/real": -0.28606855869293213, | |
| "logps/generated": -576.8764038085938, | |
| "logps/real": -475.23370361328125, | |
| "loss": 0.2603, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -1.018733263015747, | |
| "rewards/margins": 1.0629549026489258, | |
| "rewards/real": 0.044221747666597366, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 22.125, | |
| "learning_rate": 6.837606837606839e-08, | |
| "logits/generated": 0.20441779494285583, | |
| "logits/real": -0.3156498372554779, | |
| "logps/generated": -552.9418334960938, | |
| "logps/real": -434.1715393066406, | |
| "loss": 0.2664, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/generated": -1.0779039859771729, | |
| "rewards/margins": 1.1664670705795288, | |
| "rewards/real": 0.08856303989887238, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 21.75, | |
| "learning_rate": 6.481481481481481e-08, | |
| "logits/generated": 0.20344486832618713, | |
| "logits/real": -0.2555938959121704, | |
| "logps/generated": -569.4146728515625, | |
| "logps/real": -463.9098205566406, | |
| "loss": 0.2581, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -0.9887621998786926, | |
| "rewards/margins": 1.0963895320892334, | |
| "rewards/real": 0.10762734711170197, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 22.125, | |
| "learning_rate": 6.125356125356125e-08, | |
| "logits/generated": 0.23496928811073303, | |
| "logits/real": -0.2586161494255066, | |
| "logps/generated": -598.0418090820312, | |
| "logps/real": -467.5443420410156, | |
| "loss": 0.26, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -1.009071946144104, | |
| "rewards/margins": 1.0756361484527588, | |
| "rewards/real": 0.0665643960237503, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 22.5, | |
| "learning_rate": 5.7692307692307695e-08, | |
| "logits/generated": 0.25314944982528687, | |
| "logits/real": -0.2648155093193054, | |
| "logps/generated": -568.5360107421875, | |
| "logps/real": -475.3414001464844, | |
| "loss": 0.2678, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -1.0467432737350464, | |
| "rewards/margins": 1.0825011730194092, | |
| "rewards/real": 0.03575799986720085, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 22.25, | |
| "learning_rate": 5.413105413105413e-08, | |
| "logits/generated": 0.22809025645256042, | |
| "logits/real": -0.22145231068134308, | |
| "logps/generated": -562.1895141601562, | |
| "logps/real": -402.9983825683594, | |
| "loss": 0.2457, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -0.9787503480911255, | |
| "rewards/margins": 1.1118056774139404, | |
| "rewards/real": 0.13305525481700897, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 24.125, | |
| "learning_rate": 5.056980056980057e-08, | |
| "logits/generated": 0.2158997356891632, | |
| "logits/real": -0.33214133977890015, | |
| "logps/generated": -581.0538940429688, | |
| "logps/real": -487.0838928222656, | |
| "loss": 0.2588, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -1.1220922470092773, | |
| "rewards/margins": 1.1539920568466187, | |
| "rewards/real": 0.03189963847398758, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 21.75, | |
| "learning_rate": 4.7008547008547005e-08, | |
| "logits/generated": 0.24275963008403778, | |
| "logits/real": -0.3125799894332886, | |
| "logps/generated": -556.0437622070312, | |
| "logps/real": -493.3427734375, | |
| "loss": 0.265, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/generated": -1.0702166557312012, | |
| "rewards/margins": 1.044668197631836, | |
| "rewards/real": -0.02554836869239807, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 21.625, | |
| "learning_rate": 4.3447293447293445e-08, | |
| "logits/generated": 0.21104946732521057, | |
| "logits/real": -0.21389278769493103, | |
| "logps/generated": -585.9505615234375, | |
| "logps/real": -407.6512756347656, | |
| "loss": 0.2583, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -1.045369029045105, | |
| "rewards/margins": 1.1055140495300293, | |
| "rewards/real": 0.06014498323202133, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 21.75, | |
| "learning_rate": 3.9886039886039886e-08, | |
| "logits/generated": 0.17947638034820557, | |
| "logits/real": -0.3292226195335388, | |
| "logps/generated": -536.8143310546875, | |
| "logps/real": -448.7561950683594, | |
| "loss": 0.2664, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/generated": -0.9618861079216003, | |
| "rewards/margins": 1.0169373750686646, | |
| "rewards/real": 0.05505116656422615, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 20.875, | |
| "learning_rate": 3.632478632478633e-08, | |
| "logits/generated": 0.21649348735809326, | |
| "logits/real": -0.3055981993675232, | |
| "logps/generated": -581.56103515625, | |
| "logps/real": -459.6167907714844, | |
| "loss": 0.2434, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -1.0903055667877197, | |
| "rewards/margins": 1.1792160272598267, | |
| "rewards/real": 0.08891044557094574, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 21.0, | |
| "learning_rate": 3.276353276353276e-08, | |
| "logits/generated": 0.218745619058609, | |
| "logits/real": -0.3472137749195099, | |
| "logps/generated": -565.79931640625, | |
| "logps/real": -489.8983459472656, | |
| "loss": 0.2563, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -1.093998908996582, | |
| "rewards/margins": 1.1649422645568848, | |
| "rewards/real": 0.0709434524178505, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 22.375, | |
| "learning_rate": 2.92022792022792e-08, | |
| "logits/generated": 0.2501876950263977, | |
| "logits/real": -0.23009344935417175, | |
| "logps/generated": -569.3104248046875, | |
| "logps/real": -435.45599365234375, | |
| "loss": 0.242, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -1.0212743282318115, | |
| "rewards/margins": 1.118667721748352, | |
| "rewards/real": 0.09739341586828232, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 22.0, | |
| "learning_rate": 2.564102564102564e-08, | |
| "logits/generated": 0.2094719409942627, | |
| "logits/real": -0.2892032563686371, | |
| "logps/generated": -561.4392700195312, | |
| "logps/real": -422.3382263183594, | |
| "loss": 0.2481, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/generated": -1.0572887659072876, | |
| "rewards/margins": 1.1870763301849365, | |
| "rewards/real": 0.1297876536846161, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 21.125, | |
| "learning_rate": 2.2079772079772077e-08, | |
| "logits/generated": 0.22928491234779358, | |
| "logits/real": -0.27726420760154724, | |
| "logps/generated": -603.4669799804688, | |
| "logps/real": -450.5899353027344, | |
| "loss": 0.2473, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/generated": -1.1500332355499268, | |
| "rewards/margins": 1.2057290077209473, | |
| "rewards/real": 0.055695801973342896, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 22.125, | |
| "learning_rate": 1.8518518518518518e-08, | |
| "logits/generated": 0.17330752313137054, | |
| "logits/real": -0.3336513638496399, | |
| "logps/generated": -548.7659912109375, | |
| "logps/real": -464.267578125, | |
| "loss": 0.2472, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -1.0935777425765991, | |
| "rewards/margins": 1.205217957496643, | |
| "rewards/real": 0.11164028942584991, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 21.75, | |
| "learning_rate": 1.4957264957264956e-08, | |
| "logits/generated": 0.24052992463111877, | |
| "logits/real": -0.26754871010780334, | |
| "logps/generated": -572.6226806640625, | |
| "logps/real": -469.23388671875, | |
| "loss": 0.2562, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/generated": -0.9951506853103638, | |
| "rewards/margins": 1.117073655128479, | |
| "rewards/real": 0.12192295491695404, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.1396011396011397e-08, | |
| "logits/generated": 0.16937658190727234, | |
| "logits/real": -0.3556388020515442, | |
| "logps/generated": -551.8524169921875, | |
| "logps/real": -432.68316650390625, | |
| "loss": 0.2612, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/generated": -1.0066879987716675, | |
| "rewards/margins": 1.067731261253357, | |
| "rewards/real": 0.06104331091046333, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 22.625, | |
| "learning_rate": 7.834757834757834e-09, | |
| "logits/generated": 0.2987242341041565, | |
| "logits/real": -0.2289242297410965, | |
| "logps/generated": -591.0054931640625, | |
| "logps/real": -470.20758056640625, | |
| "loss": 0.2521, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/generated": -1.0826890468597412, | |
| "rewards/margins": 1.1511424779891968, | |
| "rewards/real": 0.06845370680093765, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 20.375, | |
| "learning_rate": 4.273504273504274e-09, | |
| "logits/generated": 0.2650791108608246, | |
| "logits/real": -0.2300834357738495, | |
| "logps/generated": -569.5623779296875, | |
| "logps/real": -417.6439514160156, | |
| "loss": 0.2514, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/generated": -1.0691241025924683, | |
| "rewards/margins": 1.168656826019287, | |
| "rewards/real": 0.09953261911869049, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 23.25, | |
| "learning_rate": 7.122507122507123e-10, | |
| "logits/generated": 0.22686879336833954, | |
| "logits/real": -0.32619014382362366, | |
| "logps/generated": -537.1260986328125, | |
| "logps/real": -499.892578125, | |
| "loss": 0.2663, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/generated": -0.9691260457038879, | |
| "rewards/margins": 1.0396109819412231, | |
| "rewards/real": 0.07048489153385162, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 781, | |
| "total_flos": 0.0, | |
| "train_loss": 0.31094988259020595, | |
| "train_runtime": 9866.2092, | |
| "train_samples_per_second": 5.068, | |
| "train_steps_per_second": 0.079 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 781, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |