| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.9493670886075947, | |
| "eval_steps": 1, | |
| "global_step": 156, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02531645569620253, | |
| "grad_norm": 57.114834738699926, | |
| "learning_rate": 3.125e-08, | |
| "logits/chosen": 3.3286685943603516, | |
| "logits/rejected": 3.4021615982055664, | |
| "logps/chosen": -32.21625900268555, | |
| "logps/rejected": -38.764957427978516, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05063291139240506, | |
| "grad_norm": 55.56377943215762, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": 2.647796630859375, | |
| "logits/rejected": 2.733036994934082, | |
| "logps/chosen": -24.994104385375977, | |
| "logps/rejected": -29.78559112548828, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0759493670886076, | |
| "grad_norm": 61.65658878594894, | |
| "learning_rate": 9.375e-08, | |
| "logits/chosen": 2.33656644821167, | |
| "logits/rejected": 2.518265962600708, | |
| "logps/chosen": -28.91141128540039, | |
| "logps/rejected": -35.79951477050781, | |
| "loss": 0.7163, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.046910468488931656, | |
| "rewards/margins": -0.0570392943918705, | |
| "rewards/rejected": 0.01012883335351944, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.10126582278481013, | |
| "grad_norm": 54.332498272912595, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": 2.9658162593841553, | |
| "logits/rejected": 3.0012216567993164, | |
| "logps/chosen": -29.552040100097656, | |
| "logps/rejected": -34.8863639831543, | |
| "loss": 0.7065, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.047019001096487045, | |
| "rewards/margins": -0.010130930691957474, | |
| "rewards/rejected": -0.03688807412981987, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.12658227848101267, | |
| "grad_norm": 55.132631513210136, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": 3.340583086013794, | |
| "logits/rejected": 3.325289487838745, | |
| "logps/chosen": -29.905162811279297, | |
| "logps/rejected": -38.20769500732422, | |
| "loss": 0.7069, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.06609320640563965, | |
| "rewards/margins": -0.07786056399345398, | |
| "rewards/rejected": 0.01176736131310463, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 57.59086219970169, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": 2.2882869243621826, | |
| "logits/rejected": 2.7739901542663574, | |
| "logps/chosen": -23.262107849121094, | |
| "logps/rejected": -38.5238151550293, | |
| "loss": 0.7044, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.004146900027990341, | |
| "rewards/margins": -0.010677304118871689, | |
| "rewards/rejected": 0.006530404556542635, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.17721518987341772, | |
| "grad_norm": 51.18146497947048, | |
| "learning_rate": 2.1875e-07, | |
| "logits/chosen": 2.7440547943115234, | |
| "logits/rejected": 2.5763754844665527, | |
| "logps/chosen": -29.953731536865234, | |
| "logps/rejected": -33.62974166870117, | |
| "loss": 0.7102, | |
| "rewards/accuracies": 0.375, | |
| "rewards/chosen": -0.038678932934999466, | |
| "rewards/margins": -0.042138125747442245, | |
| "rewards/rejected": 0.0034591909497976303, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 53.68918501282459, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": 3.2939491271972656, | |
| "logits/rejected": 3.003464698791504, | |
| "logps/chosen": -28.833343505859375, | |
| "logps/rejected": -23.717449188232422, | |
| "loss": 0.6967, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.013088438659906387, | |
| "rewards/margins": 0.03873240202665329, | |
| "rewards/rejected": -0.025643955916166306, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.22784810126582278, | |
| "grad_norm": 56.10725202078908, | |
| "learning_rate": 2.8125e-07, | |
| "logits/chosen": 2.243440866470337, | |
| "logits/rejected": 2.4725394248962402, | |
| "logps/chosen": -22.95738983154297, | |
| "logps/rejected": -27.134145736694336, | |
| "loss": 0.684, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.04334143549203873, | |
| "rewards/margins": 0.07354923337697983, | |
| "rewards/rejected": -0.03020780347287655, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.25316455696202533, | |
| "grad_norm": 52.307052055168896, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": 3.2282490730285645, | |
| "logits/rejected": 3.332897424697876, | |
| "logps/chosen": -26.25811004638672, | |
| "logps/rejected": -33.90528869628906, | |
| "loss": 0.69, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.05821816250681877, | |
| "rewards/margins": 0.11382907629013062, | |
| "rewards/rejected": -0.05561092123389244, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.27848101265822783, | |
| "grad_norm": 51.178561062412925, | |
| "learning_rate": 3.4375e-07, | |
| "logits/chosen": 3.0024123191833496, | |
| "logits/rejected": 2.8223249912261963, | |
| "logps/chosen": -25.597423553466797, | |
| "logps/rejected": -27.973861694335938, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.375, | |
| "rewards/chosen": -0.006550817750394344, | |
| "rewards/margins": 0.020815890282392502, | |
| "rewards/rejected": -0.02736670896410942, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 55.45441942378304, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 2.8794424533843994, | |
| "logits/rejected": 2.8852243423461914, | |
| "logps/chosen": -24.754030227661133, | |
| "logps/rejected": -37.79888916015625, | |
| "loss": 0.6747, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.028544973582029343, | |
| "rewards/margins": 0.1005675345659256, | |
| "rewards/rejected": -0.12911249697208405, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3291139240506329, | |
| "grad_norm": 53.09392363916735, | |
| "learning_rate": 4.0625e-07, | |
| "logits/chosen": 2.4571850299835205, | |
| "logits/rejected": 2.484537363052368, | |
| "logps/chosen": -25.80780601501465, | |
| "logps/rejected": -29.53820037841797, | |
| "loss": 0.6911, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.03155774995684624, | |
| "rewards/margins": 0.07387223094701767, | |
| "rewards/rejected": -0.04231448099017143, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.35443037974683544, | |
| "grad_norm": 52.83620354330049, | |
| "learning_rate": 4.375e-07, | |
| "logits/chosen": 2.509875535964966, | |
| "logits/rejected": 2.7940757274627686, | |
| "logps/chosen": -23.630701065063477, | |
| "logps/rejected": -30.94167709350586, | |
| "loss": 0.6538, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.0901968702673912, | |
| "rewards/margins": 0.12045808881521225, | |
| "rewards/rejected": -0.030261218547821045, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.379746835443038, | |
| "grad_norm": 54.43440442749398, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": 2.544405460357666, | |
| "logits/rejected": 2.4906702041625977, | |
| "logps/chosen": -31.145732879638672, | |
| "logps/rejected": -35.83900451660156, | |
| "loss": 0.6699, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.015418028458952904, | |
| "rewards/margins": 0.1095210611820221, | |
| "rewards/rejected": -0.09410304576158524, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 56.228300664828, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 2.400423049926758, | |
| "logits/rejected": 2.620924949645996, | |
| "logps/chosen": -19.128610610961914, | |
| "logps/rejected": -33.12566375732422, | |
| "loss": 0.6456, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.02211858704686165, | |
| "rewards/margins": 0.07731227576732635, | |
| "rewards/rejected": -0.0994308590888977, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.43037974683544306, | |
| "grad_norm": 55.89193538913802, | |
| "learning_rate": 4.999370587356267e-07, | |
| "logits/chosen": 2.4614853858947754, | |
| "logits/rejected": 2.2287514209747314, | |
| "logps/chosen": -30.090726852416992, | |
| "logps/rejected": -36.996307373046875, | |
| "loss": 0.6714, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.010743052698671818, | |
| "rewards/margins": 0.19052858650684357, | |
| "rewards/rejected": -0.17978551983833313, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 47.67611457430014, | |
| "learning_rate": 4.997482666353286e-07, | |
| "logits/chosen": 3.0267255306243896, | |
| "logits/rejected": 3.2108960151672363, | |
| "logps/chosen": -24.398426055908203, | |
| "logps/rejected": -36.53913879394531, | |
| "loss": 0.6385, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.06351403892040253, | |
| "rewards/margins": 0.35269594192504883, | |
| "rewards/rejected": -0.2891818583011627, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.4810126582278481, | |
| "grad_norm": 52.87018041987955, | |
| "learning_rate": 4.99433718761614e-07, | |
| "logits/chosen": 1.927664041519165, | |
| "logits/rejected": 2.412555694580078, | |
| "logps/chosen": -23.279611587524414, | |
| "logps/rejected": -35.23127746582031, | |
| "loss": 0.6107, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.028891097754240036, | |
| "rewards/margins": 0.2483980357646942, | |
| "rewards/rejected": -0.21950694918632507, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 48.602746524439674, | |
| "learning_rate": 4.989935734988097e-07, | |
| "logits/chosen": 2.6588048934936523, | |
| "logits/rejected": 2.663198232650757, | |
| "logps/chosen": -27.881851196289062, | |
| "logps/rejected": -40.07163619995117, | |
| "loss": 0.6026, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.008863937109708786, | |
| "rewards/margins": 0.33766672015190125, | |
| "rewards/rejected": -0.32880276441574097, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5316455696202531, | |
| "grad_norm": 54.16428877770723, | |
| "learning_rate": 4.984280524733107e-07, | |
| "logits/chosen": 2.961024284362793, | |
| "logits/rejected": 2.982067584991455, | |
| "logps/chosen": -33.65000915527344, | |
| "logps/rejected": -44.283145904541016, | |
| "loss": 0.6255, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.0565052255988121, | |
| "rewards/margins": 0.3312607407569885, | |
| "rewards/rejected": -0.387766033411026, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.5569620253164557, | |
| "grad_norm": 45.46343620685162, | |
| "learning_rate": 4.977374404419837e-07, | |
| "logits/chosen": 2.731034278869629, | |
| "logits/rejected": 2.8560142517089844, | |
| "logps/chosen": -28.184062957763672, | |
| "logps/rejected": -29.667505264282227, | |
| "loss": 0.5798, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.03637414053082466, | |
| "rewards/margins": 0.30975764989852905, | |
| "rewards/rejected": -0.2733834683895111, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.5822784810126582, | |
| "grad_norm": 44.12365829762132, | |
| "learning_rate": 4.969220851487844e-07, | |
| "logits/chosen": 2.936208963394165, | |
| "logits/rejected": 2.651503086090088, | |
| "logps/chosen": -33.56488037109375, | |
| "logps/rejected": -41.19189453125, | |
| "loss": 0.5772, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.06098739802837372, | |
| "rewards/margins": 0.4615795612335205, | |
| "rewards/rejected": -0.4005921483039856, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 45.017464152313536, | |
| "learning_rate": 4.959823971496574e-07, | |
| "logits/chosen": 2.6883764266967773, | |
| "logits/rejected": 2.8039445877075195, | |
| "logps/chosen": -25.476293563842773, | |
| "logps/rejected": -27.924108505249023, | |
| "loss": 0.5669, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.003826950676739216, | |
| "rewards/margins": 0.374502956867218, | |
| "rewards/rejected": -0.37067604064941406, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.6329113924050633, | |
| "grad_norm": 42.27141207917208, | |
| "learning_rate": 4.949188496058089e-07, | |
| "logits/chosen": 2.4697136878967285, | |
| "logits/rejected": 2.4843673706054688, | |
| "logps/chosen": -29.330278396606445, | |
| "logps/rejected": -34.346717834472656, | |
| "loss": 0.5643, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.015232307836413383, | |
| "rewards/margins": 0.2725878059864044, | |
| "rewards/rejected": -0.28782013058662415, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.6582278481012658, | |
| "grad_norm": 43.50860864037927, | |
| "learning_rate": 4.937319780454559e-07, | |
| "logits/chosen": 2.4729130268096924, | |
| "logits/rejected": 2.728221893310547, | |
| "logps/chosen": -15.063579559326172, | |
| "logps/rejected": -29.431615829467773, | |
| "loss": 0.5485, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.010220827534794807, | |
| "rewards/margins": 0.6472443342208862, | |
| "rewards/rejected": -0.6370234489440918, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.6835443037974683, | |
| "grad_norm": 44.10326676433438, | |
| "learning_rate": 4.924223800941717e-07, | |
| "logits/chosen": 2.506294012069702, | |
| "logits/rejected": 2.611067295074463, | |
| "logps/chosen": -27.957015991210938, | |
| "logps/rejected": -34.36131286621094, | |
| "loss": 0.5338, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.09378986805677414, | |
| "rewards/margins": 0.7992368936538696, | |
| "rewards/rejected": -0.7054470777511597, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 42.53652604653026, | |
| "learning_rate": 4.909907151739633e-07, | |
| "logits/chosen": 2.370758295059204, | |
| "logits/rejected": 2.3593924045562744, | |
| "logps/chosen": -24.23151969909668, | |
| "logps/rejected": -28.833358764648438, | |
| "loss": 0.5266, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.015818338841199875, | |
| "rewards/margins": 0.35423219203948975, | |
| "rewards/rejected": -0.3384138345718384, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.7341772151898734, | |
| "grad_norm": 43.92462963258578, | |
| "learning_rate": 4.894377041712326e-07, | |
| "logits/chosen": 2.85575532913208, | |
| "logits/rejected": 2.9239468574523926, | |
| "logps/chosen": -24.336238861083984, | |
| "logps/rejected": -32.19173049926758, | |
| "loss": 0.5788, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.1497105062007904, | |
| "rewards/margins": 0.41176801919937134, | |
| "rewards/rejected": -0.5614784955978394, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 45.52462540230979, | |
| "learning_rate": 4.877641290737883e-07, | |
| "logits/chosen": 2.577241897583008, | |
| "logits/rejected": 2.6502695083618164, | |
| "logps/chosen": -27.839168548583984, | |
| "logps/rejected": -36.543941497802734, | |
| "loss": 0.5261, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.1546507328748703, | |
| "rewards/margins": 1.282554268836975, | |
| "rewards/rejected": -1.4372049570083618, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.7848101265822784, | |
| "grad_norm": 40.84962297403327, | |
| "learning_rate": 4.859708325770919e-07, | |
| "logits/chosen": 2.821880340576172, | |
| "logits/rejected": 2.7289583683013916, | |
| "logps/chosen": -32.00937271118164, | |
| "logps/rejected": -32.9297981262207, | |
| "loss": 0.5013, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.05743463337421417, | |
| "rewards/margins": 0.7840094566345215, | |
| "rewards/rejected": -0.8414440155029297, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 42.08408504733762, | |
| "learning_rate": 4.840587176599343e-07, | |
| "logits/chosen": 2.684711456298828, | |
| "logits/rejected": 3.0184218883514404, | |
| "logps/chosen": -29.447864532470703, | |
| "logps/rejected": -45.021995544433594, | |
| "loss": 0.5126, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.03239244967699051, | |
| "rewards/margins": 1.4539411067962646, | |
| "rewards/rejected": -1.4215487241744995, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.8354430379746836, | |
| "grad_norm": 39.28461553560508, | |
| "learning_rate": 4.820287471297597e-07, | |
| "logits/chosen": 2.2539427280426025, | |
| "logits/rejected": 2.2139289379119873, | |
| "logps/chosen": -28.69852066040039, | |
| "logps/rejected": -31.549945831298828, | |
| "loss": 0.5182, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.13107189536094666, | |
| "rewards/margins": 1.036413311958313, | |
| "rewards/rejected": -1.167485237121582, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.8607594936708861, | |
| "grad_norm": 40.986833656434555, | |
| "learning_rate": 4.798819431378626e-07, | |
| "logits/chosen": 2.653698444366455, | |
| "logits/rejected": 2.910414218902588, | |
| "logps/chosen": -21.828556060791016, | |
| "logps/rejected": -38.98440170288086, | |
| "loss": 0.4764, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.08699348568916321, | |
| "rewards/margins": 0.9170618057250977, | |
| "rewards/rejected": -0.8300682306289673, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.8860759493670886, | |
| "grad_norm": 44.21046799920892, | |
| "learning_rate": 4.776193866647039e-07, | |
| "logits/chosen": 2.2878611087799072, | |
| "logits/rejected": 2.6529476642608643, | |
| "logps/chosen": -32.880462646484375, | |
| "logps/rejected": -48.58502197265625, | |
| "loss": 0.5087, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.1624097228050232, | |
| "rewards/margins": 0.6843761205673218, | |
| "rewards/rejected": -0.8467859029769897, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 37.027834114234714, | |
| "learning_rate": 4.752422169756047e-07, | |
| "logits/chosen": 1.9958323240280151, | |
| "logits/rejected": 2.1072514057159424, | |
| "logps/chosen": -25.151729583740234, | |
| "logps/rejected": -34.56462478637695, | |
| "loss": 0.448, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.035374294966459274, | |
| "rewards/margins": 0.7525378465652466, | |
| "rewards/rejected": -0.7171635627746582, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.9367088607594937, | |
| "grad_norm": 42.723620077832024, | |
| "learning_rate": 4.7275163104709194e-07, | |
| "logits/chosen": 2.223971128463745, | |
| "logits/rejected": 2.2867026329040527, | |
| "logps/chosen": -28.47149658203125, | |
| "logps/rejected": -41.30598831176758, | |
| "loss": 0.4752, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.04645688831806183, | |
| "rewards/margins": 0.7158762812614441, | |
| "rewards/rejected": -0.7623331546783447, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.9620253164556962, | |
| "grad_norm": 37.84394079738309, | |
| "learning_rate": 4.7014888296418447e-07, | |
| "logits/chosen": 2.7294678688049316, | |
| "logits/rejected": 2.7926061153411865, | |
| "logps/chosen": -27.44596290588379, | |
| "logps/rejected": -37.265743255615234, | |
| "loss": 0.4448, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.049863241612911224, | |
| "rewards/margins": 1.077908992767334, | |
| "rewards/rejected": -1.0280457735061646, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.9873417721518988, | |
| "grad_norm": 37.542515394049275, | |
| "learning_rate": 4.6743528328892384e-07, | |
| "logits/chosen": 2.9349310398101807, | |
| "logits/rejected": 2.7823617458343506, | |
| "logps/chosen": -28.974937438964844, | |
| "logps/rejected": -31.127620697021484, | |
| "loss": 0.4247, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.06356163322925568, | |
| "rewards/margins": 0.8182165622711182, | |
| "rewards/rejected": -0.7546550035476685, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.0126582278481013, | |
| "grad_norm": 34.306915155273884, | |
| "learning_rate": 4.646121984004665e-07, | |
| "logits/chosen": 2.93210768699646, | |
| "logits/rejected": 2.952573299407959, | |
| "logps/chosen": -22.667604446411133, | |
| "logps/rejected": -33.4359130859375, | |
| "loss": 0.4023, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.15648899972438812, | |
| "rewards/margins": 1.7330085039138794, | |
| "rewards/rejected": -1.5765196084976196, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.0379746835443038, | |
| "grad_norm": 34.583417240102314, | |
| "learning_rate": 4.6168104980707103e-07, | |
| "logits/chosen": 1.8353512287139893, | |
| "logits/rejected": 1.8640844821929932, | |
| "logps/chosen": -30.39456558227539, | |
| "logps/rejected": -39.79166793823242, | |
| "loss": 0.3791, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.04578995332121849, | |
| "rewards/margins": 1.216768741607666, | |
| "rewards/rejected": -1.2625586986541748, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.0632911392405062, | |
| "grad_norm": 34.082479241604844, | |
| "learning_rate": 4.5864331343032565e-07, | |
| "logits/chosen": 1.8534090518951416, | |
| "logits/rejected": 2.3674142360687256, | |
| "logps/chosen": -24.82038116455078, | |
| "logps/rejected": -36.29607009887695, | |
| "loss": 0.3826, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.017678476870059967, | |
| "rewards/margins": 1.1888235807418823, | |
| "rewards/rejected": -1.1711452007293701, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.0886075949367089, | |
| "grad_norm": 34.56856221568917, | |
| "learning_rate": 4.555005188619775e-07, | |
| "logits/chosen": 2.3347442150115967, | |
| "logits/rejected": 2.896552562713623, | |
| "logps/chosen": -27.94430923461914, | |
| "logps/rejected": -56.70208740234375, | |
| "loss": 0.3719, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.003501923754811287, | |
| "rewards/margins": 1.5741243362426758, | |
| "rewards/rejected": -1.5706223249435425, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.1139240506329113, | |
| "grad_norm": 31.370937274347394, | |
| "learning_rate": 4.5225424859373684e-07, | |
| "logits/chosen": 2.6383707523345947, | |
| "logits/rejected": 2.5485527515411377, | |
| "logps/chosen": -19.298486709594727, | |
| "logps/rejected": -25.37928009033203, | |
| "loss": 0.3744, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.22318463027477264, | |
| "rewards/margins": 0.6971991658210754, | |
| "rewards/rejected": -0.474014550447464, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.139240506329114, | |
| "grad_norm": 34.10062660077868, | |
| "learning_rate": 4.489061372204452e-07, | |
| "logits/chosen": 2.376051664352417, | |
| "logits/rejected": 2.3887717723846436, | |
| "logps/chosen": -27.828815460205078, | |
| "logps/rejected": -35.97361755371094, | |
| "loss": 0.3561, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.13765862584114075, | |
| "rewards/margins": 0.8190426826477051, | |
| "rewards/rejected": -0.6813840866088867, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.1645569620253164, | |
| "grad_norm": 34.550138211560416, | |
| "learning_rate": 4.4545787061700746e-07, | |
| "logits/chosen": 1.5758471488952637, | |
| "logits/rejected": 1.6474454402923584, | |
| "logps/chosen": -26.35508155822754, | |
| "logps/rejected": -33.66876983642578, | |
| "loss": 0.3738, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.11953898519277573, | |
| "rewards/margins": 1.4261627197265625, | |
| "rewards/rejected": -1.3066238164901733, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.189873417721519, | |
| "grad_norm": 32.97226834077475, | |
| "learning_rate": 4.4191118508950277e-07, | |
| "logits/chosen": 2.352506160736084, | |
| "logits/rejected": 2.5829250812530518, | |
| "logps/chosen": -30.464948654174805, | |
| "logps/rejected": -48.602386474609375, | |
| "loss": 0.3773, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.28972989320755005, | |
| "rewards/margins": 1.533113956451416, | |
| "rewards/rejected": -1.2433841228485107, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.2151898734177216, | |
| "grad_norm": 31.173818641433147, | |
| "learning_rate": 4.3826786650090273e-07, | |
| "logits/chosen": 2.6230416297912598, | |
| "logits/rejected": 2.9404265880584717, | |
| "logps/chosen": -26.442947387695312, | |
| "logps/rejected": -40.41877746582031, | |
| "loss": 0.3218, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.13079282641410828, | |
| "rewards/margins": 1.8220040798187256, | |
| "rewards/rejected": -1.6912113428115845, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.240506329113924, | |
| "grad_norm": 31.65330353079338, | |
| "learning_rate": 4.345297493718352e-07, | |
| "logits/chosen": 2.0961148738861084, | |
| "logits/rejected": 2.31225848197937, | |
| "logps/chosen": -28.76830291748047, | |
| "logps/rejected": -41.471527099609375, | |
| "loss": 0.324, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.17320382595062256, | |
| "rewards/margins": 1.8003501892089844, | |
| "rewards/rejected": -1.6271462440490723, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.2658227848101267, | |
| "grad_norm": 32.0517210362914, | |
| "learning_rate": 4.3069871595684787e-07, | |
| "logits/chosen": 2.477304458618164, | |
| "logits/rejected": 2.607898473739624, | |
| "logps/chosen": -25.64235496520996, | |
| "logps/rejected": -30.789039611816406, | |
| "loss": 0.3365, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.04965633898973465, | |
| "rewards/margins": 1.5173263549804688, | |
| "rewards/rejected": -1.4676700830459595, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.2911392405063291, | |
| "grad_norm": 34.75297133689012, | |
| "learning_rate": 4.2677669529663686e-07, | |
| "logits/chosen": 2.4322314262390137, | |
| "logits/rejected": 2.3749780654907227, | |
| "logps/chosen": -28.2196102142334, | |
| "logps/rejected": -36.66654968261719, | |
| "loss": 0.376, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.014162863604724407, | |
| "rewards/margins": 1.2780933380126953, | |
| "rewards/rejected": -1.292256236076355, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.3164556962025316, | |
| "grad_norm": 30.679462903169394, | |
| "learning_rate": 4.227656622467162e-07, | |
| "logits/chosen": 2.16550350189209, | |
| "logits/rejected": 2.483671188354492, | |
| "logps/chosen": -27.87743377685547, | |
| "logps/rejected": -35.79537582397461, | |
| "loss": 0.328, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.016814690083265305, | |
| "rewards/margins": 0.7990709543228149, | |
| "rewards/rejected": -0.8158857226371765, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.3417721518987342, | |
| "grad_norm": 33.362419000257866, | |
| "learning_rate": 4.186676364830186e-07, | |
| "logits/chosen": 1.6940243244171143, | |
| "logits/rejected": 1.8324474096298218, | |
| "logps/chosen": -22.575485229492188, | |
| "logps/rejected": -37.813812255859375, | |
| "loss": 0.3899, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.18489199876785278, | |
| "rewards/margins": 1.3045916557312012, | |
| "rewards/rejected": -1.4894835948944092, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.3670886075949367, | |
| "grad_norm": 34.27040807040801, | |
| "learning_rate": 4.1448468148492814e-07, | |
| "logits/chosen": 2.4027748107910156, | |
| "logits/rejected": 2.6370065212249756, | |
| "logps/chosen": -25.746681213378906, | |
| "logps/rejected": -39.262149810791016, | |
| "loss": 0.3439, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.14556269347667694, | |
| "rewards/margins": 1.5188615322113037, | |
| "rewards/rejected": -1.664424180984497, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.3924050632911391, | |
| "grad_norm": 30.395014985650786, | |
| "learning_rate": 4.10218903496256e-07, | |
| "logits/chosen": 1.9827089309692383, | |
| "logits/rejected": 2.0387563705444336, | |
| "logps/chosen": -23.929346084594727, | |
| "logps/rejected": -31.56798553466797, | |
| "loss": 0.3347, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.012682847678661346, | |
| "rewards/margins": 1.1076587438583374, | |
| "rewards/rejected": -1.0949759483337402, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.4177215189873418, | |
| "grad_norm": 30.07232120551146, | |
| "learning_rate": 4.058724504646834e-07, | |
| "logits/chosen": 2.144460916519165, | |
| "logits/rejected": 2.3883056640625, | |
| "logps/chosen": -29.003910064697266, | |
| "logps/rejected": -48.30259704589844, | |
| "loss": 0.3069, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.12483976036310196, | |
| "rewards/margins": 2.0643203258514404, | |
| "rewards/rejected": -1.9394805431365967, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.4430379746835442, | |
| "grad_norm": 29.37804816237709, | |
| "learning_rate": 4.0144751096020497e-07, | |
| "logits/chosen": 2.4025626182556152, | |
| "logits/rejected": 2.6997711658477783, | |
| "logps/chosen": -20.084449768066406, | |
| "logps/rejected": -41.12477493286133, | |
| "loss": 0.2932, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.05090276896953583, | |
| "rewards/margins": 1.5857845544815063, | |
| "rewards/rejected": -1.6366872787475586, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.4683544303797469, | |
| "grad_norm": 28.707949914480356, | |
| "learning_rate": 3.9694631307311825e-07, | |
| "logits/chosen": 1.7879891395568848, | |
| "logits/rejected": 1.9085421562194824, | |
| "logps/chosen": -24.52581214904785, | |
| "logps/rejected": -42.650333404541016, | |
| "loss": 0.2848, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.09958541393280029, | |
| "rewards/margins": 1.8922325372695923, | |
| "rewards/rejected": -1.991817831993103, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.4936708860759493, | |
| "grad_norm": 30.2249376141746, | |
| "learning_rate": 3.92371123292113e-07, | |
| "logits/chosen": 2.3484342098236084, | |
| "logits/rejected": 2.6477420330047607, | |
| "logps/chosen": -25.54952621459961, | |
| "logps/rejected": -42.36656951904297, | |
| "loss": 0.2884, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.025862978771328926, | |
| "rewards/margins": 2.4160983562469482, | |
| "rewards/rejected": -2.4419612884521484, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 29.710974044035055, | |
| "learning_rate": 3.877242453630256e-07, | |
| "logits/chosen": 2.434805154800415, | |
| "logits/rejected": 2.456228256225586, | |
| "logps/chosen": -20.521100997924805, | |
| "logps/rejected": -28.69353485107422, | |
| "loss": 0.3295, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.3444966673851013, | |
| "rewards/margins": 2.016552448272705, | |
| "rewards/rejected": -1.672055959701538, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.5443037974683544, | |
| "grad_norm": 33.402955887536194, | |
| "learning_rate": 3.8300801912883414e-07, | |
| "logits/chosen": 2.511925220489502, | |
| "logits/rejected": 2.43487286567688, | |
| "logps/chosen": -30.91568946838379, | |
| "logps/rejected": -38.48870086669922, | |
| "loss": 0.297, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.009598510339856148, | |
| "rewards/margins": 1.2743103504180908, | |
| "rewards/rejected": -1.2839089632034302, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.5696202531645569, | |
| "grad_norm": 34.221583251712744, | |
| "learning_rate": 3.7822481935147655e-07, | |
| "logits/chosen": 2.2030210494995117, | |
| "logits/rejected": 2.4195468425750732, | |
| "logps/chosen": -26.423166275024414, | |
| "logps/rejected": -36.79461669921875, | |
| "loss": 0.3281, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.2358216643333435, | |
| "rewards/margins": 1.8056204319000244, | |
| "rewards/rejected": -2.0414421558380127, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.5949367088607596, | |
| "grad_norm": 27.39193509195394, | |
| "learning_rate": 3.7337705451608667e-07, | |
| "logits/chosen": 2.442267894744873, | |
| "logits/rejected": 2.6541049480438232, | |
| "logps/chosen": -24.891948699951172, | |
| "logps/rejected": -40.318904876708984, | |
| "loss": 0.2564, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.08293318003416061, | |
| "rewards/margins": 1.7536163330078125, | |
| "rewards/rejected": -1.8365494012832642, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.620253164556962, | |
| "grad_norm": 32.95811212676831, | |
| "learning_rate": 3.6846716561824967e-07, | |
| "logits/chosen": 2.404311180114746, | |
| "logits/rejected": 2.50986385345459, | |
| "logps/chosen": -24.17348861694336, | |
| "logps/rejected": -35.517120361328125, | |
| "loss": 0.3643, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.00808901246637106, | |
| "rewards/margins": 1.9171310663223267, | |
| "rewards/rejected": -1.925220251083374, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.6455696202531644, | |
| "grad_norm": 31.89102457068134, | |
| "learning_rate": 3.634976249348867e-07, | |
| "logits/chosen": 1.8697429895401, | |
| "logits/rejected": 1.9369537830352783, | |
| "logps/chosen": -26.27237319946289, | |
| "logps/rejected": -43.584659576416016, | |
| "loss": 0.3047, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.018543783575296402, | |
| "rewards/margins": 1.540452480316162, | |
| "rewards/rejected": -1.5219086408615112, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.6708860759493671, | |
| "grad_norm": 29.55969704854734, | |
| "learning_rate": 3.584709347793895e-07, | |
| "logits/chosen": 2.12164044380188, | |
| "logits/rejected": 2.12618350982666, | |
| "logps/chosen": -19.538429260253906, | |
| "logps/rejected": -34.374759674072266, | |
| "loss": 0.3159, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.1228712797164917, | |
| "rewards/margins": 1.97737455368042, | |
| "rewards/rejected": -1.8545031547546387, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.6962025316455698, | |
| "grad_norm": 29.63374295371732, | |
| "learning_rate": 3.5338962624163016e-07, | |
| "logits/chosen": 2.3299381732940674, | |
| "logits/rejected": 2.5256855487823486, | |
| "logps/chosen": -24.86815643310547, | |
| "logps/rejected": -32.63747024536133, | |
| "loss": 0.304, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.14013400673866272, | |
| "rewards/margins": 1.9178043603897095, | |
| "rewards/rejected": -2.05793833732605, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.721518987341772, | |
| "grad_norm": 29.816270030062956, | |
| "learning_rate": 3.482562579134809e-07, | |
| "logits/chosen": 2.947877883911133, | |
| "logits/rejected": 2.919872760772705, | |
| "logps/chosen": -24.169422149658203, | |
| "logps/rejected": -35.171504974365234, | |
| "loss": 0.2947, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.18604925274848938, | |
| "rewards/margins": 2.742734432220459, | |
| "rewards/rejected": -2.556685209274292, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.7468354430379747, | |
| "grad_norm": 28.204359911989563, | |
| "learning_rate": 3.430734146004863e-07, | |
| "logits/chosen": 2.15493106842041, | |
| "logits/rejected": 2.3170015811920166, | |
| "logps/chosen": -22.2800350189209, | |
| "logps/rejected": -35.93194580078125, | |
| "loss": 0.2844, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.17889335751533508, | |
| "rewards/margins": 2.3182311058044434, | |
| "rewards/rejected": -2.497124433517456, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.7721518987341773, | |
| "grad_norm": 26.787627596958465, | |
| "learning_rate": 3.378437060203357e-07, | |
| "logits/chosen": 2.5400497913360596, | |
| "logits/rejected": 2.5072226524353027, | |
| "logps/chosen": -21.618453979492188, | |
| "logps/rejected": -33.122589111328125, | |
| "loss": 0.2766, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.05331309512257576, | |
| "rewards/margins": 2.7737677097320557, | |
| "rewards/rejected": -2.720454692840576, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.7974683544303798, | |
| "grad_norm": 31.205788121930055, | |
| "learning_rate": 3.325697654887918e-07, | |
| "logits/chosen": 1.4359736442565918, | |
| "logits/rejected": 1.0823308229446411, | |
| "logps/chosen": -26.589662551879883, | |
| "logps/rejected": -38.18647766113281, | |
| "loss": 0.2822, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.2071526050567627, | |
| "rewards/margins": 2.4934191703796387, | |
| "rewards/rejected": -2.286266565322876, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.8227848101265822, | |
| "grad_norm": 29.788415172955876, | |
| "learning_rate": 3.272542485937368e-07, | |
| "logits/chosen": 2.800675868988037, | |
| "logits/rejected": 2.9707438945770264, | |
| "logps/chosen": -25.928173065185547, | |
| "logps/rejected": -38.25574493408203, | |
| "loss": 0.2955, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0182593185454607, | |
| "rewards/margins": 2.1046228408813477, | |
| "rewards/rejected": -2.0863637924194336, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.8481012658227849, | |
| "grad_norm": 31.197362620805283, | |
| "learning_rate": 3.218998318580043e-07, | |
| "logits/chosen": 2.2215137481689453, | |
| "logits/rejected": 2.1676788330078125, | |
| "logps/chosen": -23.131969451904297, | |
| "logps/rejected": -34.484920501708984, | |
| "loss": 0.276, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.19679217040538788, | |
| "rewards/margins": 2.1039047241210938, | |
| "rewards/rejected": -1.9071124792099, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.8734177215189873, | |
| "grad_norm": 27.14206740038897, | |
| "learning_rate": 3.1650921139166874e-07, | |
| "logits/chosen": 2.0885462760925293, | |
| "logits/rejected": 2.1695892810821533, | |
| "logps/chosen": -24.332992553710938, | |
| "logps/rejected": -39.35213088989258, | |
| "loss": 0.2598, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.13339604437351227, | |
| "rewards/margins": 2.6932549476623535, | |
| "rewards/rejected": -2.559858798980713, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.8987341772151898, | |
| "grad_norm": 31.93987675909966, | |
| "learning_rate": 3.110851015344735e-07, | |
| "logits/chosen": 2.1219115257263184, | |
| "logits/rejected": 2.1965231895446777, | |
| "logps/chosen": -25.27581024169922, | |
| "logps/rejected": -34.003536224365234, | |
| "loss": 0.3142, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.1961415410041809, | |
| "rewards/margins": 2.4378857612609863, | |
| "rewards/rejected": -2.24174427986145, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.9240506329113924, | |
| "grad_norm": 29.021478328579693, | |
| "learning_rate": 3.056302334890786e-07, | |
| "logits/chosen": 1.8746654987335205, | |
| "logits/rejected": 2.389371871948242, | |
| "logps/chosen": -22.446157455444336, | |
| "logps/rejected": -39.833770751953125, | |
| "loss": 0.2671, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.18668973445892334, | |
| "rewards/margins": 3.256779909133911, | |
| "rewards/rejected": -3.0700900554656982, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.9493670886075949, | |
| "grad_norm": 31.861473137885586, | |
| "learning_rate": 3.001473539458182e-07, | |
| "logits/chosen": 2.0852279663085938, | |
| "logits/rejected": 2.2871861457824707, | |
| "logps/chosen": -27.48174285888672, | |
| "logps/rejected": -38.38823699951172, | |
| "loss": 0.2623, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.12376511096954346, | |
| "rewards/margins": 2.879195213317871, | |
| "rewards/rejected": -2.755429983139038, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.9746835443037973, | |
| "grad_norm": 30.580581160948984, | |
| "learning_rate": 2.9463922369965915e-07, | |
| "logits/chosen": 2.183192253112793, | |
| "logits/rejected": 1.8703649044036865, | |
| "logps/chosen": -25.529539108276367, | |
| "logps/rejected": -38.35371780395508, | |
| "loss": 0.2812, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.09511762112379074, | |
| "rewards/margins": 1.8313283920288086, | |
| "rewards/rejected": -1.9264459609985352, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 28.28932497860918, | |
| "learning_rate": 2.8910861626005773e-07, | |
| "logits/chosen": 2.1175851821899414, | |
| "logits/rejected": 2.7737534046173096, | |
| "logps/chosen": -22.948108673095703, | |
| "logps/rejected": -54.805484771728516, | |
| "loss": 0.2737, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.07266075909137726, | |
| "rewards/margins": 2.9845314025878906, | |
| "rewards/rejected": -3.057192087173462, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.0253164556962027, | |
| "grad_norm": 22.84435575128268, | |
| "learning_rate": 2.8355831645441387e-07, | |
| "logits/chosen": 2.129598617553711, | |
| "logits/rejected": 2.214559316635132, | |
| "logps/chosen": -33.85695266723633, | |
| "logps/rejected": -40.15086364746094, | |
| "loss": 0.2116, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04292801395058632, | |
| "rewards/margins": 2.009500741958618, | |
| "rewards/rejected": -2.052428722381592, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 22.68444850198704, | |
| "learning_rate": 2.7799111902582693e-07, | |
| "logits/chosen": 2.512767791748047, | |
| "logits/rejected": 2.5582704544067383, | |
| "logps/chosen": -25.28483009338379, | |
| "logps/rejected": -36.232017517089844, | |
| "loss": 0.2011, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.14101354777812958, | |
| "rewards/margins": 2.3776185512542725, | |
| "rewards/rejected": -2.236604690551758, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.0759493670886076, | |
| "grad_norm": 23.440024916384186, | |
| "learning_rate": 2.7240982722585837e-07, | |
| "logits/chosen": 1.941485047340393, | |
| "logits/rejected": 2.076082944869995, | |
| "logps/chosen": -25.91452407836914, | |
| "logps/rejected": -34.40084457397461, | |
| "loss": 0.2306, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.1705533266067505, | |
| "rewards/margins": 1.7886393070220947, | |
| "rewards/rejected": -1.9591926336288452, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 24.39463073368113, | |
| "learning_rate": 2.6681725140300995e-07, | |
| "logits/chosen": 1.5756129026412964, | |
| "logits/rejected": 1.9195802211761475, | |
| "logps/chosen": -21.2039737701416, | |
| "logps/rejected": -43.85890579223633, | |
| "loss": 0.2048, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.012569274753332138, | |
| "rewards/margins": 3.1985719203948975, | |
| "rewards/rejected": -3.2111411094665527, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.1265822784810124, | |
| "grad_norm": 21.14214956646525, | |
| "learning_rate": 2.6121620758762875e-07, | |
| "logits/chosen": 2.5736641883850098, | |
| "logits/rejected": 2.7494866847991943, | |
| "logps/chosen": -27.16839027404785, | |
| "logps/rejected": -37.205440521240234, | |
| "loss": 0.1933, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.010486233979463577, | |
| "rewards/margins": 2.571336269378662, | |
| "rewards/rejected": -2.560849905014038, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.151898734177215, | |
| "grad_norm": 21.59512651806526, | |
| "learning_rate": 2.5560951607395126e-07, | |
| "logits/chosen": 2.4533567428588867, | |
| "logits/rejected": 2.3579037189483643, | |
| "logps/chosen": -28.007740020751953, | |
| "logps/rejected": -36.33196258544922, | |
| "loss": 0.192, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.04641330614686012, | |
| "rewards/margins": 1.9481606483459473, | |
| "rewards/rejected": -1.901747465133667, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.1772151898734178, | |
| "grad_norm": 24.879344750048478, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": 2.1627023220062256, | |
| "logits/rejected": 2.059532403945923, | |
| "logps/chosen": -32.905269622802734, | |
| "logps/rejected": -31.235652923583984, | |
| "loss": 0.2379, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.02136234939098358, | |
| "rewards/margins": 1.5867081880569458, | |
| "rewards/rejected": -1.6080706119537354, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 22.69000503277292, | |
| "learning_rate": 2.4439048392604877e-07, | |
| "logits/chosen": 2.3072502613067627, | |
| "logits/rejected": 2.6580121517181396, | |
| "logps/chosen": -23.412647247314453, | |
| "logps/rejected": -42.0001106262207, | |
| "loss": 0.2005, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.16801214218139648, | |
| "rewards/margins": 2.332923412322998, | |
| "rewards/rejected": -2.5009355545043945, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.2278481012658227, | |
| "grad_norm": 24.693555115896494, | |
| "learning_rate": 2.3878379241237134e-07, | |
| "logits/chosen": 2.4124932289123535, | |
| "logits/rejected": 2.2819831371307373, | |
| "logps/chosen": -25.710378646850586, | |
| "logps/rejected": -32.972537994384766, | |
| "loss": 0.2211, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0028405077755451202, | |
| "rewards/margins": 2.1923816204071045, | |
| "rewards/rejected": -2.1895413398742676, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.2531645569620253, | |
| "grad_norm": 22.755579101890042, | |
| "learning_rate": 2.3318274859699008e-07, | |
| "logits/chosen": 2.006254196166992, | |
| "logits/rejected": 2.4240238666534424, | |
| "logps/chosen": -26.35239028930664, | |
| "logps/rejected": -42.78788757324219, | |
| "loss": 0.2437, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.381151020526886, | |
| "rewards/margins": 2.395555019378662, | |
| "rewards/rejected": -2.7767062187194824, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.278481012658228, | |
| "grad_norm": 23.73790882415418, | |
| "learning_rate": 2.2759017277414164e-07, | |
| "logits/chosen": 2.482501745223999, | |
| "logits/rejected": 2.456974744796753, | |
| "logps/chosen": -32.156795501708984, | |
| "logps/rejected": -35.610713958740234, | |
| "loss": 0.2328, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.08202718943357468, | |
| "rewards/margins": 1.4416414499282837, | |
| "rewards/rejected": -1.523668646812439, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 22.47212835812806, | |
| "learning_rate": 2.2200888097417302e-07, | |
| "logits/chosen": 2.5853986740112305, | |
| "logits/rejected": 2.6601390838623047, | |
| "logps/chosen": -33.39643859863281, | |
| "logps/rejected": -43.778709411621094, | |
| "loss": 0.1828, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.010579407215118408, | |
| "rewards/margins": 2.473262310028076, | |
| "rewards/rejected": -2.483841896057129, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.329113924050633, | |
| "grad_norm": 22.509363583762184, | |
| "learning_rate": 2.164416835455862e-07, | |
| "logits/chosen": 2.3972582817077637, | |
| "logits/rejected": 2.534813642501831, | |
| "logps/chosen": -22.962604522705078, | |
| "logps/rejected": -37.55415344238281, | |
| "loss": 0.2213, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.0858815610408783, | |
| "rewards/margins": 2.756964683532715, | |
| "rewards/rejected": -2.842846155166626, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 22.253217260948635, | |
| "learning_rate": 2.1089138373994222e-07, | |
| "logits/chosen": 2.153351306915283, | |
| "logits/rejected": 2.3578603267669678, | |
| "logps/chosen": -22.811275482177734, | |
| "logps/rejected": -39.044639587402344, | |
| "loss": 0.1953, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.2400507628917694, | |
| "rewards/margins": 2.045787811279297, | |
| "rewards/rejected": -2.2858386039733887, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.379746835443038, | |
| "grad_norm": 23.13742394622883, | |
| "learning_rate": 2.0536077630034085e-07, | |
| "logits/chosen": 1.7368831634521484, | |
| "logits/rejected": 1.9436362981796265, | |
| "logps/chosen": -26.176815032958984, | |
| "logps/rejected": -40.83625411987305, | |
| "loss": 0.2061, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.39153048396110535, | |
| "rewards/margins": 2.8905537128448486, | |
| "rewards/rejected": -2.499023199081421, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 21.94757471208971, | |
| "learning_rate": 1.998526460541818e-07, | |
| "logits/chosen": 2.3502449989318848, | |
| "logits/rejected": 2.3544671535491943, | |
| "logps/chosen": -26.833280563354492, | |
| "logps/rejected": -37.404605865478516, | |
| "loss": 0.1962, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.0008228495717048645, | |
| "rewards/margins": 2.2684872150421143, | |
| "rewards/rejected": -2.2676644325256348, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.430379746835443, | |
| "grad_norm": 20.124080148253146, | |
| "learning_rate": 1.9436976651092142e-07, | |
| "logits/chosen": 1.1564311981201172, | |
| "logits/rejected": 1.8700743913650513, | |
| "logps/chosen": -25.664215087890625, | |
| "logps/rejected": -42.25444030761719, | |
| "loss": 0.1803, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.013121239840984344, | |
| "rewards/margins": 2.5192627906799316, | |
| "rewards/rejected": -2.5061416625976562, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.4556962025316453, | |
| "grad_norm": 22.336736253170702, | |
| "learning_rate": 1.8891489846552644e-07, | |
| "logits/chosen": 2.3158211708068848, | |
| "logits/rejected": 2.718118667602539, | |
| "logps/chosen": -23.182619094848633, | |
| "logps/rejected": -41.77816390991211, | |
| "loss": 0.2237, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.010374765843153, | |
| "rewards/margins": 2.8450756072998047, | |
| "rewards/rejected": -2.8347012996673584, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.481012658227848, | |
| "grad_norm": 21.95104798778307, | |
| "learning_rate": 1.8349078860833124e-07, | |
| "logits/chosen": 2.2794296741485596, | |
| "logits/rejected": 2.176607847213745, | |
| "logps/chosen": -33.21621322631836, | |
| "logps/rejected": -38.940025329589844, | |
| "loss": 0.1762, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.013390496373176575, | |
| "rewards/margins": 2.8369154930114746, | |
| "rewards/rejected": -2.8503060340881348, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 22.129805040414322, | |
| "learning_rate": 1.781001681419957e-07, | |
| "logits/chosen": 1.8799982070922852, | |
| "logits/rejected": 2.1679162979125977, | |
| "logps/chosen": -20.39841079711914, | |
| "logps/rejected": -40.38523864746094, | |
| "loss": 0.2032, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.04029037803411484, | |
| "rewards/margins": 2.2781383991241455, | |
| "rewards/rejected": -2.2378478050231934, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.5316455696202533, | |
| "grad_norm": 18.828604860640993, | |
| "learning_rate": 1.7274575140626315e-07, | |
| "logits/chosen": 2.4270401000976562, | |
| "logits/rejected": 2.768435478210449, | |
| "logps/chosen": -28.433048248291016, | |
| "logps/rejected": -47.266197204589844, | |
| "loss": 0.1502, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.2632801830768585, | |
| "rewards/margins": 2.7442736625671387, | |
| "rewards/rejected": -3.007554054260254, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.5569620253164556, | |
| "grad_norm": 20.880074717956166, | |
| "learning_rate": 1.674302345112083e-07, | |
| "logits/chosen": 1.9880956411361694, | |
| "logits/rejected": 2.2008445262908936, | |
| "logps/chosen": -24.262271881103516, | |
| "logps/rejected": -45.60854721069336, | |
| "loss": 0.2089, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.04222775995731354, | |
| "rewards/margins": 3.01908802986145, | |
| "rewards/rejected": -2.9768600463867188, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.5822784810126582, | |
| "grad_norm": 21.838535925363054, | |
| "learning_rate": 1.621562939796643e-07, | |
| "logits/chosen": 2.205430746078491, | |
| "logits/rejected": 2.4456026554107666, | |
| "logps/chosen": -24.633563995361328, | |
| "logps/rejected": -37.809478759765625, | |
| "loss": 0.2187, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.2103780210018158, | |
| "rewards/margins": 2.4951250553131104, | |
| "rewards/rejected": -2.2847468852996826, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 22.277589914673204, | |
| "learning_rate": 1.569265853995137e-07, | |
| "logits/chosen": 1.4377330541610718, | |
| "logits/rejected": 1.4871635437011719, | |
| "logps/chosen": -24.512638092041016, | |
| "logps/rejected": -37.595733642578125, | |
| "loss": 0.1964, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.3236038088798523, | |
| "rewards/margins": 2.621075391769409, | |
| "rewards/rejected": -2.9446792602539062, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.632911392405063, | |
| "grad_norm": 21.07481872439906, | |
| "learning_rate": 1.517437420865191e-07, | |
| "logits/chosen": 1.6525717973709106, | |
| "logits/rejected": 2.0419435501098633, | |
| "logps/chosen": -24.287813186645508, | |
| "logps/rejected": -43.26752853393555, | |
| "loss": 0.1873, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.2492070347070694, | |
| "rewards/margins": 3.5699872970581055, | |
| "rewards/rejected": -3.3207802772521973, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 20.677444088444823, | |
| "learning_rate": 1.4661037375836987e-07, | |
| "logits/chosen": 2.6604502201080322, | |
| "logits/rejected": 2.587705373764038, | |
| "logps/chosen": -25.429487228393555, | |
| "logps/rejected": -36.87110137939453, | |
| "loss": 0.1855, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.07871632277965546, | |
| "rewards/margins": 2.1279850006103516, | |
| "rewards/rejected": -2.2067012786865234, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.6835443037974684, | |
| "grad_norm": 23.5751426913938, | |
| "learning_rate": 1.4152906522061047e-07, | |
| "logits/chosen": 2.1078615188598633, | |
| "logits/rejected": 2.089411735534668, | |
| "logps/chosen": -25.450345993041992, | |
| "logps/rejected": -33.47675323486328, | |
| "loss": 0.1944, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.11998379230499268, | |
| "rewards/margins": 2.037046432495117, | |
| "rewards/rejected": -1.917062759399414, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 22.767643882863975, | |
| "learning_rate": 1.365023750651133e-07, | |
| "logits/chosen": 1.7534555196762085, | |
| "logits/rejected": 2.2760560512542725, | |
| "logps/chosen": -27.695234298706055, | |
| "logps/rejected": -46.436767578125, | |
| "loss": 0.1638, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.030735721811652184, | |
| "rewards/margins": 3.136977434158325, | |
| "rewards/rejected": -3.1677134037017822, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.7341772151898733, | |
| "grad_norm": 20.58426575505033, | |
| "learning_rate": 1.3153283438175034e-07, | |
| "logits/chosen": 2.3597166538238525, | |
| "logits/rejected": 2.5851235389709473, | |
| "logps/chosen": -23.936431884765625, | |
| "logps/rejected": -47.1025390625, | |
| "loss": 0.1785, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.1647489368915558, | |
| "rewards/margins": 3.3675107955932617, | |
| "rewards/rejected": -3.532259941101074, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.759493670886076, | |
| "grad_norm": 21.349116101829402, | |
| "learning_rate": 1.2662294548391328e-07, | |
| "logits/chosen": 1.9581003189086914, | |
| "logits/rejected": 1.994449496269226, | |
| "logps/chosen": -28.851770401000977, | |
| "logps/rejected": -39.374019622802734, | |
| "loss": 0.1772, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.023888343945145607, | |
| "rewards/margins": 2.033907890319824, | |
| "rewards/rejected": -2.0100197792053223, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.7848101265822782, | |
| "grad_norm": 20.523600951939002, | |
| "learning_rate": 1.2177518064852348e-07, | |
| "logits/chosen": 2.149280071258545, | |
| "logits/rejected": 2.338773012161255, | |
| "logps/chosen": -24.151710510253906, | |
| "logps/rejected": -44.250370025634766, | |
| "loss": 0.197, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008578717708587646, | |
| "rewards/margins": 2.208198308944702, | |
| "rewards/rejected": -2.1996195316314697, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 21.793053109415474, | |
| "learning_rate": 1.1699198087116588e-07, | |
| "logits/chosen": 1.9670238494873047, | |
| "logits/rejected": 2.2674317359924316, | |
| "logps/chosen": -19.423011779785156, | |
| "logps/rejected": -43.466705322265625, | |
| "loss": 0.1912, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.0942595824599266, | |
| "rewards/margins": 2.482712984085083, | |
| "rewards/rejected": -2.576972484588623, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.8354430379746836, | |
| "grad_norm": 18.7267582468967, | |
| "learning_rate": 1.1227575463697439e-07, | |
| "logits/chosen": 2.125753402709961, | |
| "logits/rejected": 2.3489747047424316, | |
| "logps/chosen": -24.716629028320312, | |
| "logps/rejected": -35.32937240600586, | |
| "loss": 0.1497, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.06388971209526062, | |
| "rewards/margins": 2.4911532402038574, | |
| "rewards/rejected": -2.5550432205200195, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.8607594936708862, | |
| "grad_norm": 21.953907512144703, | |
| "learning_rate": 1.0762887670788701e-07, | |
| "logits/chosen": 1.75847327709198, | |
| "logits/rejected": 1.766452431678772, | |
| "logps/chosen": -33.99777603149414, | |
| "logps/rejected": -49.58311080932617, | |
| "loss": 0.1887, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.09381230175495148, | |
| "rewards/margins": 3.5883729457855225, | |
| "rewards/rejected": -3.682185649871826, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.8860759493670884, | |
| "grad_norm": 21.263544038283367, | |
| "learning_rate": 1.0305368692688174e-07, | |
| "logits/chosen": 1.9139025211334229, | |
| "logits/rejected": 1.8650161027908325, | |
| "logps/chosen": -17.992523193359375, | |
| "logps/rejected": -38.388519287109375, | |
| "loss": 0.1854, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.09737363457679749, | |
| "rewards/margins": 3.4405698776245117, | |
| "rewards/rejected": -3.3431966304779053, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 19.070922554256914, | |
| "learning_rate": 9.855248903979505e-08, | |
| "logits/chosen": 2.087536096572876, | |
| "logits/rejected": 2.118912935256958, | |
| "logps/chosen": -30.79559326171875, | |
| "logps/rejected": -47.3471565246582, | |
| "loss": 0.1693, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.07975523173809052, | |
| "rewards/margins": 3.129413366317749, | |
| "rewards/rejected": -3.2091684341430664, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.9367088607594938, | |
| "grad_norm": 21.46961371171142, | |
| "learning_rate": 9.412754953531663e-08, | |
| "logits/chosen": 1.7869651317596436, | |
| "logits/rejected": 1.881705641746521, | |
| "logps/chosen": -24.043615341186523, | |
| "logps/rejected": -38.01569366455078, | |
| "loss": 0.1852, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.10924415290355682, | |
| "rewards/margins": 2.661862373352051, | |
| "rewards/rejected": -2.5526182651519775, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 19.929759669575105, | |
| "learning_rate": 8.978109650374396e-08, | |
| "logits/chosen": 1.7048578262329102, | |
| "logits/rejected": 1.9161901473999023, | |
| "logps/chosen": -26.87795639038086, | |
| "logps/rejected": -42.6606330871582, | |
| "loss": 0.1589, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.13115011155605316, | |
| "rewards/margins": 3.121182441711426, | |
| "rewards/rejected": -2.990032196044922, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.9873417721518987, | |
| "grad_norm": 25.118856703851232, | |
| "learning_rate": 8.551531851507185e-08, | |
| "logits/chosen": 2.3607137203216553, | |
| "logits/rejected": 2.5181257724761963, | |
| "logps/chosen": -30.176773071289062, | |
| "logps/rejected": -44.625003814697266, | |
| "loss": 0.224, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.30477192997932434, | |
| "rewards/margins": 2.4585647583007812, | |
| "rewards/rejected": -2.763336658477783, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.0126582278481013, | |
| "grad_norm": 24.215057475486347, | |
| "learning_rate": 8.133236351698142e-08, | |
| "logits/chosen": 1.6108707189559937, | |
| "logits/rejected": 1.7527028322219849, | |
| "logps/chosen": -18.345888137817383, | |
| "logps/rejected": -42.66941833496094, | |
| "loss": 0.2033, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02492370456457138, | |
| "rewards/margins": 3.3315927982330322, | |
| "rewards/rejected": -3.306669235229492, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.037974683544304, | |
| "grad_norm": 18.940438234622814, | |
| "learning_rate": 7.723433775328384e-08, | |
| "logits/chosen": 2.0556139945983887, | |
| "logits/rejected": 2.3014981746673584, | |
| "logps/chosen": -21.665407180786133, | |
| "logps/rejected": -41.777896881103516, | |
| "loss": 0.1836, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.17341355979442596, | |
| "rewards/margins": 2.674577236175537, | |
| "rewards/rejected": -2.5011634826660156, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.0632911392405062, | |
| "grad_norm": 18.297883681094888, | |
| "learning_rate": 7.322330470336313e-08, | |
| "logits/chosen": 2.301410675048828, | |
| "logits/rejected": 2.2454984188079834, | |
| "logps/chosen": -23.815502166748047, | |
| "logps/rejected": -41.840091705322266, | |
| "loss": 0.1505, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02547144889831543, | |
| "rewards/margins": 3.206540584564209, | |
| "rewards/rejected": -3.1810688972473145, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.088607594936709, | |
| "grad_norm": 18.507140827349733, | |
| "learning_rate": 6.930128404315214e-08, | |
| "logits/chosen": 2.0005404949188232, | |
| "logits/rejected": 2.2419371604919434, | |
| "logps/chosen": -23.189014434814453, | |
| "logps/rejected": -44.778133392333984, | |
| "loss": 0.1697, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.21837717294692993, | |
| "rewards/margins": 2.497182846069336, | |
| "rewards/rejected": -2.715559959411621, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.1139240506329116, | |
| "grad_norm": 20.458085524449526, | |
| "learning_rate": 6.547025062816486e-08, | |
| "logits/chosen": 2.5110156536102295, | |
| "logits/rejected": 2.835902452468872, | |
| "logps/chosen": -21.34715461730957, | |
| "logps/rejected": -42.06149673461914, | |
| "loss": 0.1769, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.2778359651565552, | |
| "rewards/margins": 3.1256825923919678, | |
| "rewards/rejected": -2.847846508026123, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.1392405063291138, | |
| "grad_norm": 19.657986671859284, | |
| "learning_rate": 6.173213349909728e-08, | |
| "logits/chosen": 2.4499258995056152, | |
| "logits/rejected": 2.4804461002349854, | |
| "logps/chosen": -24.912944793701172, | |
| "logps/rejected": -38.67371368408203, | |
| "loss": 0.172, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.06854649633169174, | |
| "rewards/margins": 2.4865663051605225, | |
| "rewards/rejected": -2.5551130771636963, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.1645569620253164, | |
| "grad_norm": 18.87674548749799, | |
| "learning_rate": 5.808881491049722e-08, | |
| "logits/chosen": 1.3738226890563965, | |
| "logits/rejected": 1.6148171424865723, | |
| "logps/chosen": -25.101566314697266, | |
| "logps/rejected": -40.51878356933594, | |
| "loss": 0.1786, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.33920496702194214, | |
| "rewards/margins": 2.1792898178100586, | |
| "rewards/rejected": -2.5184948444366455, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.189873417721519, | |
| "grad_norm": 20.113483471649342, | |
| "learning_rate": 5.454212938299255e-08, | |
| "logits/chosen": 2.0692553520202637, | |
| "logits/rejected": 2.1005959510803223, | |
| "logps/chosen": -26.763628005981445, | |
| "logps/rejected": -37.69628143310547, | |
| "loss": 0.1624, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04393099620938301, | |
| "rewards/margins": 2.375842809677124, | |
| "rewards/rejected": -2.419773817062378, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.2151898734177213, | |
| "grad_norm": 19.33193488722329, | |
| "learning_rate": 5.109386277955477e-08, | |
| "logits/chosen": 1.9354777336120605, | |
| "logits/rejected": 2.3222734928131104, | |
| "logps/chosen": -22.444690704345703, | |
| "logps/rejected": -42.732322692871094, | |
| "loss": 0.1615, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.1692490577697754, | |
| "rewards/margins": 2.950117349624634, | |
| "rewards/rejected": -2.7808682918548584, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.240506329113924, | |
| "grad_norm": 17.657011863002797, | |
| "learning_rate": 4.774575140626316e-08, | |
| "logits/chosen": 1.7825671434402466, | |
| "logits/rejected": 2.0206806659698486, | |
| "logps/chosen": -26.314062118530273, | |
| "logps/rejected": -44.959896087646484, | |
| "loss": 0.1292, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.18020519614219666, | |
| "rewards/margins": 3.4240498542785645, | |
| "rewards/rejected": -3.243844747543335, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.2658227848101267, | |
| "grad_norm": 17.36688426753045, | |
| "learning_rate": 4.449948113802254e-08, | |
| "logits/chosen": 1.963316559791565, | |
| "logits/rejected": 2.073986530303955, | |
| "logps/chosen": -29.986591339111328, | |
| "logps/rejected": -46.73766326904297, | |
| "loss": 0.1476, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.1175435334444046, | |
| "rewards/margins": 3.3601157665252686, | |
| "rewards/rejected": -3.477659225463867, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.291139240506329, | |
| "grad_norm": 17.66376104354918, | |
| "learning_rate": 4.1356686569674335e-08, | |
| "logits/chosen": 1.4700018167495728, | |
| "logits/rejected": 1.7809115648269653, | |
| "logps/chosen": -27.53030014038086, | |
| "logps/rejected": -43.54850387573242, | |
| "loss": 0.1642, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.1137082502245903, | |
| "rewards/margins": 2.9248647689819336, | |
| "rewards/rejected": -2.811156749725342, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.3164556962025316, | |
| "grad_norm": 19.797305830292228, | |
| "learning_rate": 3.831895019292897e-08, | |
| "logits/chosen": 1.8914071321487427, | |
| "logits/rejected": 1.8628309965133667, | |
| "logps/chosen": -25.410337448120117, | |
| "logps/rejected": -42.06156921386719, | |
| "loss": 0.1525, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.044954825192689896, | |
| "rewards/margins": 2.8561882972717285, | |
| "rewards/rejected": -2.9011433124542236, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 3.3417721518987342, | |
| "grad_norm": 18.851413557062596, | |
| "learning_rate": 3.538780159953347e-08, | |
| "logits/chosen": 2.0251123905181885, | |
| "logits/rejected": 2.210965156555176, | |
| "logps/chosen": -20.657623291015625, | |
| "logps/rejected": -45.97637939453125, | |
| "loss": 0.1488, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.21129505336284637, | |
| "rewards/margins": 3.6020712852478027, | |
| "rewards/rejected": -3.3907763957977295, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 3.367088607594937, | |
| "grad_norm": 18.383406639630245, | |
| "learning_rate": 3.256471671107616e-08, | |
| "logits/chosen": 1.5468469858169556, | |
| "logits/rejected": 1.8942471742630005, | |
| "logps/chosen": -34.759246826171875, | |
| "logps/rejected": -54.94669723510742, | |
| "loss": 0.161, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.08364612609148026, | |
| "rewards/margins": 3.432290554046631, | |
| "rewards/rejected": -3.348644495010376, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 3.392405063291139, | |
| "grad_norm": 18.442661691122066, | |
| "learning_rate": 2.98511170358155e-08, | |
| "logits/chosen": 1.9410444498062134, | |
| "logits/rejected": 1.9713021516799927, | |
| "logps/chosen": -33.468505859375, | |
| "logps/rejected": -40.28277587890625, | |
| "loss": 0.1517, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.1791667342185974, | |
| "rewards/margins": 2.8069636821746826, | |
| "rewards/rejected": -2.986130475997925, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 3.4177215189873418, | |
| "grad_norm": 19.123016466489233, | |
| "learning_rate": 2.724836895290805e-08, | |
| "logits/chosen": 2.172494649887085, | |
| "logits/rejected": 2.3771188259124756, | |
| "logps/chosen": -18.799209594726562, | |
| "logps/rejected": -27.414329528808594, | |
| "loss": 0.1574, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.05184905230998993, | |
| "rewards/margins": 1.7008121013641357, | |
| "rewards/rejected": -1.648963212966919, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 3.4430379746835444, | |
| "grad_norm": 19.389448097902036, | |
| "learning_rate": 2.475778302439524e-08, | |
| "logits/chosen": 1.5435059070587158, | |
| "logits/rejected": 2.0342299938201904, | |
| "logps/chosen": -23.09898567199707, | |
| "logps/rejected": -41.35944366455078, | |
| "loss": 0.1673, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.01025979220867157, | |
| "rewards/margins": 3.0842089653015137, | |
| "rewards/rejected": -3.094468832015991, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 3.4683544303797467, | |
| "grad_norm": 18.168150870329598, | |
| "learning_rate": 2.2380613335296033e-08, | |
| "logits/chosen": 1.5661985874176025, | |
| "logits/rejected": 1.6321345567703247, | |
| "logps/chosen": -21.557811737060547, | |
| "logps/rejected": -37.444087982177734, | |
| "loss": 0.1457, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.32843151688575745, | |
| "rewards/margins": 3.217196464538574, | |
| "rewards/rejected": -2.888765335083008, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 3.4936708860759493, | |
| "grad_norm": 18.10048373008008, | |
| "learning_rate": 2.0118056862137354e-08, | |
| "logits/chosen": 2.2138497829437256, | |
| "logits/rejected": 2.3827362060546875, | |
| "logps/chosen": -25.050493240356445, | |
| "logps/rejected": -40.5790901184082, | |
| "loss": 0.1574, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.05445308983325958, | |
| "rewards/margins": 2.6866414546966553, | |
| "rewards/rejected": -2.7410943508148193, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 3.518987341772152, | |
| "grad_norm": 17.844608513376855, | |
| "learning_rate": 1.797125287024029e-08, | |
| "logits/chosen": 2.252528667449951, | |
| "logits/rejected": 2.517775297164917, | |
| "logps/chosen": -25.718944549560547, | |
| "logps/rejected": -40.703636169433594, | |
| "loss": 0.1633, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.23676270246505737, | |
| "rewards/margins": 2.2277328968048096, | |
| "rewards/rejected": -2.4644956588745117, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 3.5443037974683547, | |
| "grad_norm": 19.82721673855706, | |
| "learning_rate": 1.5941282340065697e-08, | |
| "logits/chosen": 2.1378254890441895, | |
| "logits/rejected": 2.3837673664093018, | |
| "logps/chosen": -21.759735107421875, | |
| "logps/rejected": -35.555789947509766, | |
| "loss": 0.1725, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.32154813408851624, | |
| "rewards/margins": 2.5137619972229004, | |
| "rewards/rejected": -2.192214012145996, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.569620253164557, | |
| "grad_norm": 19.687984330411137, | |
| "learning_rate": 1.4029167422908105e-08, | |
| "logits/chosen": 1.8691318035125732, | |
| "logits/rejected": 1.951183557510376, | |
| "logps/chosen": -22.21922492980957, | |
| "logps/rejected": -38.082176208496094, | |
| "loss": 0.1587, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.12926092743873596, | |
| "rewards/margins": 3.5831055641174316, | |
| "rewards/rejected": -3.4538450241088867, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 3.5949367088607596, | |
| "grad_norm": 19.210992875859024, | |
| "learning_rate": 1.2235870926211616e-08, | |
| "logits/chosen": 1.878167748451233, | |
| "logits/rejected": 2.1185381412506104, | |
| "logps/chosen": -25.917736053466797, | |
| "logps/rejected": -45.47472381591797, | |
| "loss": 0.1359, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0160946324467659, | |
| "rewards/margins": 3.645671844482422, | |
| "rewards/rejected": -3.629577398300171, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 3.620253164556962, | |
| "grad_norm": 20.116883252259765, | |
| "learning_rate": 1.0562295828767387e-08, | |
| "logits/chosen": 1.2207794189453125, | |
| "logits/rejected": 1.1344801187515259, | |
| "logps/chosen": -23.36668586730957, | |
| "logps/rejected": -31.341829299926758, | |
| "loss": 0.1841, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.16667915880680084, | |
| "rewards/margins": 1.5721771717071533, | |
| "rewards/rejected": -1.7388561964035034, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 3.6455696202531644, | |
| "grad_norm": 19.720241233310634, | |
| "learning_rate": 9.009284826036689e-09, | |
| "logits/chosen": 2.4453952312469482, | |
| "logits/rejected": 2.662018299102783, | |
| "logps/chosen": -22.387048721313477, | |
| "logps/rejected": -36.92119216918945, | |
| "loss": 0.1617, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.031851306557655334, | |
| "rewards/margins": 2.5965640544891357, | |
| "rewards/rejected": -2.5647125244140625, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.670886075949367, | |
| "grad_norm": 17.86661138635714, | |
| "learning_rate": 7.577619905828281e-09, | |
| "logits/chosen": 2.745617628097534, | |
| "logits/rejected": 2.844001531600952, | |
| "logps/chosen": -26.597410202026367, | |
| "logps/rejected": -36.82072448730469, | |
| "loss": 0.142, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.17600667476654053, | |
| "rewards/margins": 2.553907632827759, | |
| "rewards/rejected": -2.7299141883850098, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.6962025316455698, | |
| "grad_norm": 16.0477696571449, | |
| "learning_rate": 6.268021954544095e-09, | |
| "logits/chosen": 2.1746973991394043, | |
| "logits/rejected": 2.2613565921783447, | |
| "logps/chosen": -32.3970947265625, | |
| "logps/rejected": -40.7056884765625, | |
| "loss": 0.1372, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.09012872725725174, | |
| "rewards/margins": 4.017242431640625, | |
| "rewards/rejected": -4.1073713302612305, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 3.721518987341772, | |
| "grad_norm": 18.112204362761403, | |
| "learning_rate": 5.08115039419113e-09, | |
| "logits/chosen": 2.2588939666748047, | |
| "logits/rejected": 2.217837333679199, | |
| "logps/chosen": -23.21999740600586, | |
| "logps/rejected": -38.22819519042969, | |
| "loss": 0.1574, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.07647465169429779, | |
| "rewards/margins": 3.384047031402588, | |
| "rewards/rejected": -3.460521697998047, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 3.7468354430379747, | |
| "grad_norm": 19.146595468250382, | |
| "learning_rate": 4.0176028503425826e-09, | |
| "logits/chosen": 2.2898073196411133, | |
| "logits/rejected": 2.612645149230957, | |
| "logps/chosen": -21.40774917602539, | |
| "logps/rejected": -39.644142150878906, | |
| "loss": 0.1575, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.14267057180404663, | |
| "rewards/margins": 2.3787121772766113, | |
| "rewards/rejected": -2.23604154586792, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 3.7721518987341773, | |
| "grad_norm": 16.066181981705437, | |
| "learning_rate": 3.077914851215585e-09, | |
| "logits/chosen": 2.0213255882263184, | |
| "logits/rejected": 2.22232723236084, | |
| "logps/chosen": -20.719486236572266, | |
| "logps/rejected": -33.82309341430664, | |
| "loss": 0.1448, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04364713281393051, | |
| "rewards/margins": 2.438537836074829, | |
| "rewards/rejected": -2.482184886932373, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 3.7974683544303796, | |
| "grad_norm": 20.168906914662397, | |
| "learning_rate": 2.2625595580163247e-09, | |
| "logits/chosen": 1.8664175271987915, | |
| "logits/rejected": 2.135633945465088, | |
| "logps/chosen": -24.14510726928711, | |
| "logps/rejected": -39.59516906738281, | |
| "loss": 0.1844, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.004626031965017319, | |
| "rewards/margins": 2.253840446472168, | |
| "rewards/rejected": -2.2584664821624756, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.8227848101265822, | |
| "grad_norm": 21.962898348700495, | |
| "learning_rate": 1.5719475266893489e-09, | |
| "logits/chosen": 1.7901225090026855, | |
| "logits/rejected": 1.9218878746032715, | |
| "logps/chosen": -22.082956314086914, | |
| "logps/rejected": -43.000667572021484, | |
| "loss": 0.192, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.07910831272602081, | |
| "rewards/margins": 2.6802215576171875, | |
| "rewards/rejected": -2.6011130809783936, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 3.848101265822785, | |
| "grad_norm": 16.511033653339577, | |
| "learning_rate": 1.0064265011902328e-09, | |
| "logits/chosen": 2.4072799682617188, | |
| "logits/rejected": 2.4377200603485107, | |
| "logps/chosen": -25.803665161132812, | |
| "logps/rejected": -40.92607879638672, | |
| "loss": 0.1549, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.035745151340961456, | |
| "rewards/margins": 3.1956663131713867, | |
| "rewards/rejected": -3.159921169281006, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 3.8734177215189876, | |
| "grad_norm": 18.636452488465395, | |
| "learning_rate": 5.662812383859794e-10, | |
| "logits/chosen": 2.361091136932373, | |
| "logits/rejected": 2.3107473850250244, | |
| "logps/chosen": -27.91613006591797, | |
| "logps/rejected": -39.84597396850586, | |
| "loss": 0.1544, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.1522400677204132, | |
| "rewards/margins": 3.036912441253662, | |
| "rewards/rejected": -3.189152240753174, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 3.8987341772151898, | |
| "grad_norm": 17.763028400462222, | |
| "learning_rate": 2.5173336467135263e-10, | |
| "logits/chosen": 2.0824742317199707, | |
| "logits/rejected": 2.3250155448913574, | |
| "logps/chosen": -23.416114807128906, | |
| "logps/rejected": -38.985511779785156, | |
| "loss": 0.1542, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.06766586005687714, | |
| "rewards/margins": 2.1996846199035645, | |
| "rewards/rejected": -2.267350673675537, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 3.9240506329113924, | |
| "grad_norm": 19.457837238018815, | |
| "learning_rate": 6.294126437336733e-11, | |
| "logits/chosen": 2.00872540473938, | |
| "logits/rejected": 2.4769511222839355, | |
| "logps/chosen": -24.827199935913086, | |
| "logps/rejected": -44.62668991088867, | |
| "loss": 0.1614, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.19776426255702972, | |
| "rewards/margins": 2.720571756362915, | |
| "rewards/rejected": -2.9183361530303955, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.9493670886075947, | |
| "grad_norm": 19.565085246341297, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 1.8402495384216309, | |
| "logits/rejected": 2.033839464187622, | |
| "logps/chosen": -28.689044952392578, | |
| "logps/rejected": -46.74669647216797, | |
| "loss": 0.1636, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.04416623339056969, | |
| "rewards/margins": 3.22959041595459, | |
| "rewards/rejected": -3.185424327850342, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.9493670886075947, | |
| "step": 156, | |
| "total_flos": 0.0, | |
| "train_loss": 0.3210896818110576, | |
| "train_runtime": 45702.9058, | |
| "train_samples_per_second": 0.442, | |
| "train_steps_per_second": 0.003 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 156, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 200, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |