| { | |
| "best_metric": 0.6840614080429077, | |
| "best_model_checkpoint": "./output/checkpoints/2024-05-27_09-01-43/checkpoint-100", | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 1271, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003933910306845004, | |
| "grad_norm": 33.997314453125, | |
| "learning_rate": 1.25e-06, | |
| "logits/chosen": -0.2329253852367401, | |
| "logits/rejected": -0.7133080363273621, | |
| "logps/chosen": -206.9918670654297, | |
| "logps/rejected": -177.71676635742188, | |
| "loss": 0.6965, | |
| "rewards/accuracies": 0.16249999403953552, | |
| "rewards/chosen": -0.0024444584269076586, | |
| "rewards/margins": -0.0065179443918168545, | |
| "rewards/rejected": 0.004073486663401127, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.007867820613690008, | |
| "grad_norm": 33.46182632446289, | |
| "learning_rate": 2.8125e-06, | |
| "logits/chosen": -0.39770540595054626, | |
| "logits/rejected": -0.7366135120391846, | |
| "logps/chosen": -200.96145629882812, | |
| "logps/rejected": -177.04241943359375, | |
| "loss": 0.6857, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.028690528124570847, | |
| "rewards/margins": 0.01567094773054123, | |
| "rewards/rejected": 0.013019581325352192, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.011801730920535013, | |
| "grad_norm": 36.08251953125, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "logits/chosen": -0.35487601161003113, | |
| "logits/rejected": -0.6588561534881592, | |
| "logps/chosen": -217.158203125, | |
| "logps/rejected": -193.87928771972656, | |
| "loss": 0.6847, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.1420166790485382, | |
| "rewards/margins": 0.021750029176473618, | |
| "rewards/rejected": 0.12026665359735489, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.015735641227380016, | |
| "grad_norm": 31.359071731567383, | |
| "learning_rate": 5.9375e-06, | |
| "logits/chosen": -0.38902169466018677, | |
| "logits/rejected": -0.7328646779060364, | |
| "logps/chosen": -208.9059600830078, | |
| "logps/rejected": -179.52752685546875, | |
| "loss": 0.6636, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.23876996338367462, | |
| "rewards/margins": 0.07814005762338638, | |
| "rewards/rejected": 0.16062989830970764, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01966955153422502, | |
| "grad_norm": 37.41947555541992, | |
| "learning_rate": 7.500000000000001e-06, | |
| "logits/chosen": -0.24571442604064941, | |
| "logits/rejected": -0.7017894983291626, | |
| "logps/chosen": -195.68539428710938, | |
| "logps/rejected": -178.2770233154297, | |
| "loss": 0.6703, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.38804444670677185, | |
| "rewards/margins": 0.09177973121404648, | |
| "rewards/rejected": 0.2962647080421448, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.023603461841070025, | |
| "grad_norm": 36.168155670166016, | |
| "learning_rate": 9.0625e-06, | |
| "logits/chosen": -0.3279009461402893, | |
| "logits/rejected": -0.62000572681427, | |
| "logps/chosen": -207.6711883544922, | |
| "logps/rejected": -183.57412719726562, | |
| "loss": 0.6945, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.5336718559265137, | |
| "rewards/margins": 0.08061937242746353, | |
| "rewards/rejected": 0.45305246114730835, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02753737214791503, | |
| "grad_norm": 33.33737564086914, | |
| "learning_rate": 1.0625e-05, | |
| "logits/chosen": -0.4809085428714752, | |
| "logits/rejected": -0.758043646812439, | |
| "logps/chosen": -220.7525634765625, | |
| "logps/rejected": -198.58531188964844, | |
| "loss": 0.6564, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.8875558972358704, | |
| "rewards/margins": 0.20188398659229279, | |
| "rewards/rejected": 0.6856719255447388, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03147128245476003, | |
| "grad_norm": 32.628173828125, | |
| "learning_rate": 1.2187500000000001e-05, | |
| "logits/chosen": -0.34948527812957764, | |
| "logits/rejected": -0.7154799699783325, | |
| "logps/chosen": -195.3294677734375, | |
| "logps/rejected": -178.76821899414062, | |
| "loss": 0.6995, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.8487609624862671, | |
| "rewards/margins": 0.11841963231563568, | |
| "rewards/rejected": 0.7303413152694702, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03540519276160504, | |
| "grad_norm": 34.517208099365234, | |
| "learning_rate": 1.3125e-05, | |
| "logits/chosen": -0.6799963712692261, | |
| "logits/rejected": -1.0354492664337158, | |
| "logps/chosen": -196.0514373779297, | |
| "logps/rejected": -165.4368438720703, | |
| "loss": 0.6874, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.6915052533149719, | |
| "rewards/margins": 0.18386030197143555, | |
| "rewards/rejected": 0.5076450109481812, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03933910306845004, | |
| "grad_norm": 39.48639678955078, | |
| "learning_rate": 1.4687500000000001e-05, | |
| "logits/chosen": -0.22616450488567352, | |
| "logits/rejected": -0.3378121256828308, | |
| "logps/chosen": -210.38534545898438, | |
| "logps/rejected": -205.64596557617188, | |
| "loss": 0.6612, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.6935254335403442, | |
| "rewards/margins": 0.17534935474395752, | |
| "rewards/rejected": 0.5181760191917419, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.043273013375295044, | |
| "grad_norm": 54.90278244018555, | |
| "learning_rate": 1.6250000000000002e-05, | |
| "logits/chosen": -0.5758259892463684, | |
| "logits/rejected": -0.7866605520248413, | |
| "logps/chosen": -194.25332641601562, | |
| "logps/rejected": -179.38743591308594, | |
| "loss": 0.7132, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.7601491212844849, | |
| "rewards/margins": 0.13087329268455505, | |
| "rewards/rejected": 0.629275918006897, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04720692368214005, | |
| "grad_norm": 31.500804901123047, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "logits/chosen": -0.3127598762512207, | |
| "logits/rejected": -0.7857316136360168, | |
| "logps/chosen": -214.54507446289062, | |
| "logps/rejected": -167.3846435546875, | |
| "loss": 0.659, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.8511220216751099, | |
| "rewards/margins": 0.3019554316997528, | |
| "rewards/rejected": 0.5491665601730347, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05114083398898505, | |
| "grad_norm": 39.18369674682617, | |
| "learning_rate": 1.9062500000000003e-05, | |
| "logits/chosen": -0.4257170557975769, | |
| "logits/rejected": -0.7206593751907349, | |
| "logps/chosen": -212.1534881591797, | |
| "logps/rejected": -198.5198516845703, | |
| "loss": 0.6603, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 1.0609362125396729, | |
| "rewards/margins": 0.4154728055000305, | |
| "rewards/rejected": 0.6454635858535767, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05507474429583006, | |
| "grad_norm": 23.921350479125977, | |
| "learning_rate": 2.0625000000000003e-05, | |
| "logits/chosen": -0.4161883294582367, | |
| "logits/rejected": -0.7778038382530212, | |
| "logps/chosen": -211.8742218017578, | |
| "logps/rejected": -173.7216033935547, | |
| "loss": 0.6306, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.1546392440795898, | |
| "rewards/margins": 0.4130212366580963, | |
| "rewards/rejected": 0.7416179180145264, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.059008654602675056, | |
| "grad_norm": 33.804962158203125, | |
| "learning_rate": 2.21875e-05, | |
| "logits/chosen": -0.11207835376262665, | |
| "logits/rejected": -0.5466287136077881, | |
| "logps/chosen": -206.02218627929688, | |
| "logps/rejected": -185.80941772460938, | |
| "loss": 0.719, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.1464534997940063, | |
| "rewards/margins": 0.2297494113445282, | |
| "rewards/rejected": 0.9167040586471558, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06294256490952006, | |
| "grad_norm": 42.39493942260742, | |
| "learning_rate": 2.375e-05, | |
| "logits/chosen": -0.2344416081905365, | |
| "logits/rejected": -0.49077630043029785, | |
| "logps/chosen": -209.4600067138672, | |
| "logps/rejected": -198.47911071777344, | |
| "loss": 0.663, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.7387064099311829, | |
| "rewards/margins": 0.3134633004665375, | |
| "rewards/rejected": 0.4252430498600006, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06687647521636507, | |
| "grad_norm": 35.38875961303711, | |
| "learning_rate": 2.5312500000000002e-05, | |
| "logits/chosen": -0.26731210947036743, | |
| "logits/rejected": -0.7659153342247009, | |
| "logps/chosen": -206.7689666748047, | |
| "logps/rejected": -178.71986389160156, | |
| "loss": 0.7082, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.22556963562965393, | |
| "rewards/margins": 0.251675546169281, | |
| "rewards/rejected": -0.026105916127562523, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07081038552321008, | |
| "grad_norm": 30.606552124023438, | |
| "learning_rate": 2.6875000000000003e-05, | |
| "logits/chosen": -0.34513598680496216, | |
| "logits/rejected": -0.5674473643302917, | |
| "logps/chosen": -208.49002075195312, | |
| "logps/rejected": -184.6549072265625, | |
| "loss": 0.7702, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.2557719945907593, | |
| "rewards/margins": 0.22060665488243103, | |
| "rewards/rejected": 0.03516533225774765, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07474429583005507, | |
| "grad_norm": 25.326486587524414, | |
| "learning_rate": 2.8437500000000003e-05, | |
| "logits/chosen": -0.21977896988391876, | |
| "logits/rejected": -0.6439090371131897, | |
| "logps/chosen": -201.21401977539062, | |
| "logps/rejected": -183.35049438476562, | |
| "loss": 0.713, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.6074048280715942, | |
| "rewards/margins": 0.23941688239574432, | |
| "rewards/rejected": 0.3679879307746887, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07867820613690008, | |
| "grad_norm": 34.922325134277344, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "logits/chosen": -0.4304371774196625, | |
| "logits/rejected": -0.7793359756469727, | |
| "logps/chosen": -207.411376953125, | |
| "logps/rejected": -178.08729553222656, | |
| "loss": 0.7514, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.6943355798721313, | |
| "rewards/margins": 0.1873859167098999, | |
| "rewards/rejected": 0.5069497227668762, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07867820613690008, | |
| "eval_logits/chosen": 1.2959624528884888, | |
| "eval_logits/rejected": 1.0611025094985962, | |
| "eval_logps/chosen": -205.19613647460938, | |
| "eval_logps/rejected": -178.5214385986328, | |
| "eval_loss": 0.6840614080429077, | |
| "eval_rewards/accuracies": 0.628125011920929, | |
| "eval_rewards/chosen": 1.1096659898757935, | |
| "eval_rewards/margins": 0.36510738730430603, | |
| "eval_rewards/rejected": 0.7445584535598755, | |
| "eval_runtime": 248.7873, | |
| "eval_samples_per_second": 2.572, | |
| "eval_steps_per_second": 0.161, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08261211644374508, | |
| "grad_norm": 27.505342483520508, | |
| "learning_rate": 3.15625e-05, | |
| "logits/chosen": -0.19241487979888916, | |
| "logits/rejected": -0.49540799856185913, | |
| "logps/chosen": -200.0576629638672, | |
| "logps/rejected": -175.6849365234375, | |
| "loss": 0.6671, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 1.2072901725769043, | |
| "rewards/margins": 0.31199443340301514, | |
| "rewards/rejected": 0.8952957391738892, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.08654602675059009, | |
| "grad_norm": 34.53056335449219, | |
| "learning_rate": 3.3125000000000006e-05, | |
| "logits/chosen": -0.4692454934120178, | |
| "logits/rejected": -0.9546957015991211, | |
| "logps/chosen": -194.42092895507812, | |
| "logps/rejected": -168.50645446777344, | |
| "loss": 0.6457, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 1.2674047946929932, | |
| "rewards/margins": 0.4998513162136078, | |
| "rewards/rejected": 0.767553448677063, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0904799370574351, | |
| "grad_norm": 35.058509826660156, | |
| "learning_rate": 3.46875e-05, | |
| "logits/chosen": -0.6155041456222534, | |
| "logits/rejected": -0.9042506217956543, | |
| "logps/chosen": -198.44775390625, | |
| "logps/rejected": -174.51780700683594, | |
| "loss": 0.8259, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.7558386921882629, | |
| "rewards/margins": 0.17122629284858704, | |
| "rewards/rejected": 0.5846124291419983, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0944138473642801, | |
| "grad_norm": 25.631980895996094, | |
| "learning_rate": 3.625e-05, | |
| "logits/chosen": -0.546207070350647, | |
| "logits/rejected": -0.7782138586044312, | |
| "logps/chosen": -193.2004852294922, | |
| "logps/rejected": -179.29165649414062, | |
| "loss": 0.7377, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.7277377247810364, | |
| "rewards/margins": 0.21684861183166504, | |
| "rewards/rejected": 0.5108891725540161, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0983477576711251, | |
| "grad_norm": 33.057899475097656, | |
| "learning_rate": 3.7812500000000004e-05, | |
| "logits/chosen": -0.4421041011810303, | |
| "logits/rejected": -0.6752287745475769, | |
| "logps/chosen": -186.85108947753906, | |
| "logps/rejected": -176.91525268554688, | |
| "loss": 0.588, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.4784621298313141, | |
| "rewards/margins": 0.6664348840713501, | |
| "rewards/rejected": -0.18797270953655243, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1022816679779701, | |
| "grad_norm": 42.00132751464844, | |
| "learning_rate": 3.9375e-05, | |
| "logits/chosen": -0.3549075722694397, | |
| "logits/rejected": -0.7709188461303711, | |
| "logps/chosen": -202.1655731201172, | |
| "logps/rejected": -174.2749786376953, | |
| "loss": 0.8219, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.28374719619750977, | |
| "rewards/margins": 0.11102640628814697, | |
| "rewards/rejected": -0.3947736322879791, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10621557828481511, | |
| "grad_norm": 27.35382843017578, | |
| "learning_rate": 3.9999320096411495e-05, | |
| "logits/chosen": -0.427701473236084, | |
| "logits/rejected": -0.8705617189407349, | |
| "logps/chosen": -200.99496459960938, | |
| "logps/rejected": -165.5473175048828, | |
| "loss": 0.6218, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.44559532403945923, | |
| "rewards/margins": 0.4697811007499695, | |
| "rewards/rejected": -0.024185750633478165, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.11014948859166011, | |
| "grad_norm": 35.097049713134766, | |
| "learning_rate": 3.999516529744215e-05, | |
| "logits/chosen": -0.27036961913108826, | |
| "logits/rejected": -0.6928902864456177, | |
| "logps/chosen": -196.0111083984375, | |
| "logps/rejected": -173.6074981689453, | |
| "loss": 0.6374, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.8343067169189453, | |
| "rewards/margins": 0.6672177314758301, | |
| "rewards/rejected": 1.1670891046524048, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11408339889850512, | |
| "grad_norm": 25.420122146606445, | |
| "learning_rate": 3.99872342074443e-05, | |
| "logits/chosen": -0.21814267337322235, | |
| "logits/rejected": -0.6547086834907532, | |
| "logps/chosen": -200.3138427734375, | |
| "logps/rejected": -175.4734344482422, | |
| "loss": 0.6219, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 2.1070187091827393, | |
| "rewards/margins": 0.7536414861679077, | |
| "rewards/rejected": 1.353377103805542, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.11801730920535011, | |
| "grad_norm": 28.114505767822266, | |
| "learning_rate": 3.997552832428523e-05, | |
| "logits/chosen": -0.36429575085639954, | |
| "logits/rejected": -0.547380805015564, | |
| "logps/chosen": -190.30709838867188, | |
| "logps/rejected": -177.975830078125, | |
| "loss": 0.8049, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.508959174156189, | |
| "rewards/margins": 0.5520061254501343, | |
| "rewards/rejected": 0.9569530487060547, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12195121951219512, | |
| "grad_norm": 28.442045211791992, | |
| "learning_rate": 3.9960049858740445e-05, | |
| "logits/chosen": -0.5695027709007263, | |
| "logits/rejected": -0.7303559184074402, | |
| "logps/chosen": -196.6102294921875, | |
| "logps/rejected": -179.8834991455078, | |
| "loss": 0.8199, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.8943918943405151, | |
| "rewards/margins": 0.4739474356174469, | |
| "rewards/rejected": 0.4204444885253906, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.12588512981904013, | |
| "grad_norm": 32.67585754394531, | |
| "learning_rate": 3.994080173407612e-05, | |
| "logits/chosen": -0.3450685143470764, | |
| "logits/rejected": -0.8491863012313843, | |
| "logps/chosen": -191.87008666992188, | |
| "logps/rejected": -165.56228637695312, | |
| "loss": 0.7242, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.7810487747192383, | |
| "rewards/margins": 0.4842549264431, | |
| "rewards/rejected": 0.2967938780784607, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.12981904012588513, | |
| "grad_norm": 23.827922821044922, | |
| "learning_rate": 3.9917787585497054e-05, | |
| "logits/chosen": -0.1322830617427826, | |
| "logits/rejected": -0.6103258728981018, | |
| "logps/chosen": -212.47152709960938, | |
| "logps/rejected": -184.1310272216797, | |
| "loss": 0.6944, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.2753359079360962, | |
| "rewards/margins": 0.585142970085144, | |
| "rewards/rejected": 0.6901928186416626, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.13375295043273014, | |
| "grad_norm": 26.69400405883789, | |
| "learning_rate": 3.9891011759460066e-05, | |
| "logits/chosen": -0.2469785213470459, | |
| "logits/rejected": -0.737481415271759, | |
| "logps/chosen": -202.321044921875, | |
| "logps/rejected": -170.80929565429688, | |
| "loss": 0.7278, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 2.2151436805725098, | |
| "rewards/margins": 0.6948539018630981, | |
| "rewards/rejected": 1.5202900171279907, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.13768686073957515, | |
| "grad_norm": 37.72239303588867, | |
| "learning_rate": 3.986047931285316e-05, | |
| "logits/chosen": -0.25732505321502686, | |
| "logits/rejected": -0.6134179830551147, | |
| "logps/chosen": -215.3037109375, | |
| "logps/rejected": -183.12831115722656, | |
| "loss": 1.0019, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 3.008638620376587, | |
| "rewards/margins": 0.3702520728111267, | |
| "rewards/rejected": 2.6383864879608154, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.14162077104642015, | |
| "grad_norm": 19.040857315063477, | |
| "learning_rate": 3.982619601204049e-05, | |
| "logits/chosen": -0.0029738754965364933, | |
| "logits/rejected": -0.3499515652656555, | |
| "logps/chosen": -224.81015014648438, | |
| "logps/rejected": -198.2837371826172, | |
| "loss": 0.9635, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 2.651772975921631, | |
| "rewards/margins": 0.27305322885513306, | |
| "rewards/rejected": 2.3787198066711426, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14555468135326516, | |
| "grad_norm": 27.078737258911133, | |
| "learning_rate": 3.97881683317733e-05, | |
| "logits/chosen": -0.062339358031749725, | |
| "logits/rejected": -0.5432006120681763, | |
| "logps/chosen": -198.5349578857422, | |
| "logps/rejected": -165.4753875732422, | |
| "loss": 0.7487, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 2.0952491760253906, | |
| "rewards/margins": 0.6280597448348999, | |
| "rewards/rejected": 1.4671893119812012, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.14948859166011014, | |
| "grad_norm": 21.261680603027344, | |
| "learning_rate": 3.974640345396709e-05, | |
| "logits/chosen": -0.04105158522725105, | |
| "logits/rejected": -0.4839680790901184, | |
| "logps/chosen": -202.52001953125, | |
| "logps/rejected": -180.2872314453125, | |
| "loss": 0.7641, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 1.8239288330078125, | |
| "rewards/margins": 0.5607953667640686, | |
| "rewards/rejected": 1.2631335258483887, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.15342250196695514, | |
| "grad_norm": 24.164844512939453, | |
| "learning_rate": 3.9700909266345264e-05, | |
| "logits/chosen": -0.09652426838874817, | |
| "logits/rejected": -0.45777568221092224, | |
| "logps/chosen": -194.02333068847656, | |
| "logps/rejected": -163.8848876953125, | |
| "loss": 0.7631, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7619401216506958, | |
| "rewards/margins": 0.377643883228302, | |
| "rewards/rejected": 0.3842962384223938, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.15735641227380015, | |
| "grad_norm": 23.005035400390625, | |
| "learning_rate": 3.965169436094947e-05, | |
| "logits/chosen": -0.07853743433952332, | |
| "logits/rejected": -0.3859809637069702, | |
| "logps/chosen": -202.27133178710938, | |
| "logps/rejected": -177.63619995117188, | |
| "loss": 0.699, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.7786092162132263, | |
| "rewards/margins": 0.5468564629554749, | |
| "rewards/rejected": 0.23175275325775146, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15735641227380015, | |
| "eval_logits/chosen": 1.3732103109359741, | |
| "eval_logits/rejected": 1.1526342630386353, | |
| "eval_logps/chosen": -205.0836944580078, | |
| "eval_logps/rejected": -178.6134033203125, | |
| "eval_loss": 0.7704851627349854, | |
| "eval_rewards/accuracies": 0.609375, | |
| "eval_rewards/chosen": 1.1546350717544556, | |
| "eval_rewards/margins": 0.44686493277549744, | |
| "eval_rewards/rejected": 0.7077701687812805, | |
| "eval_runtime": 249.1371, | |
| "eval_samples_per_second": 2.569, | |
| "eval_steps_per_second": 0.161, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16129032258064516, | |
| "grad_norm": 17.80652618408203, | |
| "learning_rate": 3.959876803251684e-05, | |
| "logits/chosen": -0.01523427665233612, | |
| "logits/rejected": -0.35804808139801025, | |
| "logps/chosen": -195.39199829101562, | |
| "logps/rejected": -177.60208129882812, | |
| "loss": 0.7971, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 1.1401153802871704, | |
| "rewards/margins": 0.43174582719802856, | |
| "rewards/rejected": 0.7083694338798523, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.16522423288749016, | |
| "grad_norm": 33.28690719604492, | |
| "learning_rate": 3.954214027672465e-05, | |
| "logits/chosen": -0.0755038633942604, | |
| "logits/rejected": -0.3410184979438782, | |
| "logps/chosen": -213.752685546875, | |
| "logps/rejected": -188.33914184570312, | |
| "loss": 0.9044, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 1.4741712808609009, | |
| "rewards/margins": 0.14965666830539703, | |
| "rewards/rejected": 1.3245145082473755, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.16915814319433517, | |
| "grad_norm": 26.59926986694336, | |
| "learning_rate": 3.948182178830249e-05, | |
| "logits/chosen": 0.1018071174621582, | |
| "logits/rejected": -0.31446847319602966, | |
| "logps/chosen": -207.9669647216797, | |
| "logps/rejected": -177.61013793945312, | |
| "loss": 0.7243, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.8158692121505737, | |
| "rewards/margins": 0.5528920292854309, | |
| "rewards/rejected": 0.2629771828651428, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.17309205350118018, | |
| "grad_norm": 38.70561981201172, | |
| "learning_rate": 3.9417823959012495e-05, | |
| "logits/chosen": 0.043111931532621384, | |
| "logits/rejected": -0.27513837814331055, | |
| "logps/chosen": -207.97213745117188, | |
| "logps/rejected": -176.2123565673828, | |
| "loss": 0.8073, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 1.1955599784851074, | |
| "rewards/margins": 0.5336702466011047, | |
| "rewards/rejected": 0.6618898510932922, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.17702596380802518, | |
| "grad_norm": 24.805295944213867, | |
| "learning_rate": 3.935015887549786e-05, | |
| "logits/chosen": -0.08715621381998062, | |
| "logits/rejected": -0.451333612203598, | |
| "logps/chosen": -216.1869659423828, | |
| "logps/rejected": -184.5681915283203, | |
| "loss": 0.7354, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 1.5138676166534424, | |
| "rewards/margins": 0.6545284986495972, | |
| "rewards/rejected": 0.8593391180038452, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1809598741148702, | |
| "grad_norm": 32.434017181396484, | |
| "learning_rate": 3.9278839317000164e-05, | |
| "logits/chosen": -0.22302675247192383, | |
| "logits/rejected": -0.6187790632247925, | |
| "logps/chosen": -205.05984497070312, | |
| "logps/rejected": -178.08538818359375, | |
| "loss": 0.7056, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 1.2798973321914673, | |
| "rewards/margins": 0.8485145568847656, | |
| "rewards/rejected": 0.43138280510902405, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1848937844217152, | |
| "grad_norm": 25.297792434692383, | |
| "learning_rate": 3.920387875294588e-05, | |
| "logits/chosen": -0.14879488945007324, | |
| "logits/rejected": -0.7074145078659058, | |
| "logps/chosen": -198.54901123046875, | |
| "logps/rejected": -169.9258575439453, | |
| "loss": 0.6727, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7288762331008911, | |
| "rewards/margins": 0.656203031539917, | |
| "rewards/rejected": 0.07267318665981293, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.1888276947285602, | |
| "grad_norm": 24.259140014648438, | |
| "learning_rate": 3.9125291340402557e-05, | |
| "logits/chosen": 0.05777154117822647, | |
| "logits/rejected": -0.37030404806137085, | |
| "logps/chosen": -195.25747680664062, | |
| "logps/rejected": -167.06149291992188, | |
| "loss": 0.6126, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.6011714935302734, | |
| "rewards/margins": 0.7886489629745483, | |
| "rewards/rejected": 0.8125225305557251, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.19276160503540518, | |
| "grad_norm": 26.536211013793945, | |
| "learning_rate": 3.904309192140507e-05, | |
| "logits/chosen": -0.27970609068870544, | |
| "logits/rejected": -0.5693933367729187, | |
| "logps/chosen": -204.64598083496094, | |
| "logps/rejected": -196.45191955566406, | |
| "loss": 0.6255, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.9763765335083008, | |
| "rewards/margins": 0.9035207033157349, | |
| "rewards/rejected": 1.072855830192566, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.1966955153422502, | |
| "grad_norm": 23.359954833984375, | |
| "learning_rate": 3.89572960201526e-05, | |
| "logits/chosen": -0.007777350954711437, | |
| "logits/rejected": -0.2527710497379303, | |
| "logps/chosen": -201.5716094970703, | |
| "logps/rejected": -187.55841064453125, | |
| "loss": 0.7931, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 1.4763085842132568, | |
| "rewards/margins": 0.5630350708961487, | |
| "rewards/rejected": 0.9132736325263977, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2006294256490952, | |
| "grad_norm": 20.35169792175293, | |
| "learning_rate": 3.886791984007669e-05, | |
| "logits/chosen": -0.13616272807121277, | |
| "logits/rejected": -0.46024090051651, | |
| "logps/chosen": -218.16427612304688, | |
| "logps/rejected": -192.9780731201172, | |
| "loss": 0.5993, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.0198808908462524, | |
| "rewards/margins": 0.9105831384658813, | |
| "rewards/rejected": 0.1092977523803711, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2045633359559402, | |
| "grad_norm": 24.485294342041016, | |
| "learning_rate": 3.877498026078108e-05, | |
| "logits/chosen": -0.41903620958328247, | |
| "logits/rejected": -0.6284810304641724, | |
| "logps/chosen": -197.4494171142578, | |
| "logps/rejected": -182.7037811279297, | |
| "loss": 0.7597, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.005517137236893177, | |
| "rewards/margins": 0.6833702325820923, | |
| "rewards/rejected": -0.6778531074523926, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2084972462627852, | |
| "grad_norm": 19.309831619262695, | |
| "learning_rate": 3.867849483485383e-05, | |
| "logits/chosen": -0.4443703591823578, | |
| "logits/rejected": -0.8326670527458191, | |
| "logps/chosen": -199.72280883789062, | |
| "logps/rejected": -168.3777313232422, | |
| "loss": 0.6804, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.8990485072135925, | |
| "rewards/margins": 0.7167578935623169, | |
| "rewards/rejected": 0.18229058384895325, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.21243115656963021, | |
| "grad_norm": 20.631038665771484, | |
| "learning_rate": 3.857848178455231e-05, | |
| "logits/chosen": -0.16152219474315643, | |
| "logits/rejected": -0.6074092984199524, | |
| "logps/chosen": -212.79232788085938, | |
| "logps/rejected": -169.90603637695312, | |
| "loss": 0.6141, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.5287070274353027, | |
| "rewards/margins": 1.119511365890503, | |
| "rewards/rejected": 1.4091956615447998, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.21636506687647522, | |
| "grad_norm": 23.65202522277832, | |
| "learning_rate": 3.8474959998361754e-05, | |
| "logits/chosen": -0.3141329884529114, | |
| "logits/rejected": -0.6160004734992981, | |
| "logps/chosen": -187.09658813476562, | |
| "logps/rejected": -175.7742919921875, | |
| "loss": 0.9405, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 2.6424171924591064, | |
| "rewards/margins": 0.35450419783592224, | |
| "rewards/rejected": 2.2879130840301514, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.22029897718332023, | |
| "grad_norm": 22.0035343170166, | |
| "learning_rate": 3.8367949027427994e-05, | |
| "logits/chosen": -0.03003566339612007, | |
| "logits/rejected": -0.3478749394416809, | |
| "logps/chosen": -195.31460571289062, | |
| "logps/rejected": -176.53146362304688, | |
| "loss": 0.7537, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 3.3984057903289795, | |
| "rewards/margins": 0.9712641835212708, | |
| "rewards/rejected": 2.4271416664123535, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.22423288749016523, | |
| "grad_norm": 27.25768280029297, | |
| "learning_rate": 3.825746908186499e-05, | |
| "logits/chosen": -0.33755970001220703, | |
| "logits/rejected": -0.5320338010787964, | |
| "logps/chosen": -194.16282653808594, | |
| "logps/rejected": -177.7003631591797, | |
| "loss": 0.9776, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 2.651137351989746, | |
| "rewards/margins": 0.3439365029335022, | |
| "rewards/rejected": 2.3072009086608887, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.22816679779701024, | |
| "grad_norm": 28.454999923706055, | |
| "learning_rate": 3.8143541026937976e-05, | |
| "logits/chosen": -0.22913965582847595, | |
| "logits/rejected": -0.4301827549934387, | |
| "logps/chosen": -191.88259887695312, | |
| "logps/rejected": -186.07785034179688, | |
| "loss": 0.7314, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.7435264587402344, | |
| "rewards/margins": 0.5647959113121033, | |
| "rewards/rejected": 1.1787304878234863, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.23210070810385522, | |
| "grad_norm": 17.058061599731445, | |
| "learning_rate": 3.802618637912282e-05, | |
| "logits/chosen": -0.4652739465236664, | |
| "logits/rejected": -0.7998835444450378, | |
| "logps/chosen": -193.7137908935547, | |
| "logps/rejected": -167.9613800048828, | |
| "loss": 0.8402, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.0257607698440552, | |
| "rewards/margins": 0.507733166217804, | |
| "rewards/rejected": 0.5180276036262512, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.23603461841070023, | |
| "grad_norm": 19.008708953857422, | |
| "learning_rate": 3.7905427302042454e-05, | |
| "logits/chosen": -0.18881992995738983, | |
| "logits/rejected": -0.5596441030502319, | |
| "logps/chosen": -198.87579345703125, | |
| "logps/rejected": -176.9473876953125, | |
| "loss": 0.7099, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.2355034351348877, | |
| "rewards/margins": 0.7009516358375549, | |
| "rewards/rejected": 0.5345517992973328, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.23603461841070023, | |
| "eval_logits/chosen": 1.17328941822052, | |
| "eval_logits/rejected": 0.9396275281906128, | |
| "eval_logps/chosen": -205.6018524169922, | |
| "eval_logps/rejected": -179.09677124023438, | |
| "eval_loss": 0.8744672536849976, | |
| "eval_rewards/accuracies": 0.5703125, | |
| "eval_rewards/chosen": 0.9473776817321777, | |
| "eval_rewards/margins": 0.43294864892959595, | |
| "eval_rewards/rejected": 0.5144290924072266, | |
| "eval_runtime": 248.0407, | |
| "eval_samples_per_second": 2.58, | |
| "eval_steps_per_second": 0.161, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.23996852871754523, | |
| "grad_norm": 32.76321029663086, | |
| "learning_rate": 3.778128660228097e-05, | |
| "logits/chosen": -0.5831426382064819, | |
| "logits/rejected": -0.8806711435317993, | |
| "logps/chosen": -196.5940704345703, | |
| "logps/rejected": -165.324462890625, | |
| "loss": 0.7365, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.9338918924331665, | |
| "rewards/margins": 0.5459949374198914, | |
| "rewards/rejected": 0.3878970444202423, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.24390243902439024, | |
| "grad_norm": 25.2067813873291, | |
| "learning_rate": 3.7653787725076466e-05, | |
| "logits/chosen": -0.31566348671913147, | |
| "logits/rejected": -0.571528434753418, | |
| "logps/chosen": -206.60708618164062, | |
| "logps/rejected": -191.65855407714844, | |
| "loss": 1.0555, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 1.3235902786254883, | |
| "rewards/margins": 0.047803860157728195, | |
| "rewards/rejected": 1.2757863998413086, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.24783634933123525, | |
| "grad_norm": 21.43438148498535, | |
| "learning_rate": 3.7522954749893095e-05, | |
| "logits/chosen": -0.18173213303089142, | |
| "logits/rejected": -0.5252254605293274, | |
| "logps/chosen": -209.19760131835938, | |
| "logps/rejected": -184.69784545898438, | |
| "loss": 0.9501, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 1.7247976064682007, | |
| "rewards/margins": 0.3406931459903717, | |
| "rewards/rejected": 1.3841044902801514, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.25177025963808025, | |
| "grad_norm": 24.763315200805664, | |
| "learning_rate": 3.738881238587344e-05, | |
| "logits/chosen": -0.06764630228281021, | |
| "logits/rejected": -0.5149588584899902, | |
| "logps/chosen": -213.1123504638672, | |
| "logps/rejected": -178.50633239746094, | |
| "loss": 0.6909, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.758510947227478, | |
| "rewards/margins": 0.698074460029602, | |
| "rewards/rejected": 1.060436487197876, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.25570416994492523, | |
| "grad_norm": 29.972768783569336, | |
| "learning_rate": 3.725138596717195e-05, | |
| "logits/chosen": -0.37508150935173035, | |
| "logits/rejected": -0.9795374870300293, | |
| "logps/chosen": -210.3339385986328, | |
| "logps/rejected": -166.22647094726562, | |
| "loss": 0.8229, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.5866848230361938, | |
| "rewards/margins": 0.4931170344352722, | |
| "rewards/rejected": 1.0935678482055664, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.25963808025177026, | |
| "grad_norm": 30.266515731811523, | |
| "learning_rate": 3.711070144817032e-05, | |
| "logits/chosen": -0.5352300405502319, | |
| "logits/rejected": -0.7394737005233765, | |
| "logps/chosen": -208.7587890625, | |
| "logps/rejected": -191.28744506835938, | |
| "loss": 0.8837, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 2.188896656036377, | |
| "rewards/margins": 0.5729233026504517, | |
| "rewards/rejected": 1.6159734725952148, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.26357199055861524, | |
| "grad_norm": 20.564146041870117, | |
| "learning_rate": 3.6966785398575716e-05, | |
| "logits/chosen": -0.11968457698822021, | |
| "logits/rejected": -0.622562050819397, | |
| "logps/chosen": -217.9984130859375, | |
| "logps/rejected": -174.47152709960938, | |
| "loss": 0.6454, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 2.778174877166748, | |
| "rewards/margins": 1.0597846508026123, | |
| "rewards/rejected": 1.7183904647827148, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.2675059008654603, | |
| "grad_norm": 28.211387634277344, | |
| "learning_rate": 3.681966499840286e-05, | |
| "logits/chosen": -0.3878430426120758, | |
| "logits/rejected": -0.6015113592147827, | |
| "logps/chosen": -192.3523406982422, | |
| "logps/rejected": -186.26373291015625, | |
| "loss": 0.7849, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 2.3973546028137207, | |
| "rewards/margins": 0.8487040400505066, | |
| "rewards/rejected": 1.5486505031585693, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.27143981117230526, | |
| "grad_norm": 25.53026580810547, | |
| "learning_rate": 3.6669368032840766e-05, | |
| "logits/chosen": -0.14762678742408752, | |
| "logits/rejected": -0.5588310956954956, | |
| "logps/chosen": -228.94393920898438, | |
| "logps/rejected": -188.67474365234375, | |
| "loss": 0.7247, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 2.564577579498291, | |
| "rewards/margins": 0.7130967974662781, | |
| "rewards/rejected": 1.8514807224273682, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.2753737214791503, | |
| "grad_norm": 18.148210525512695, | |
| "learning_rate": 3.651592288700525e-05, | |
| "logits/chosen": -0.3907206058502197, | |
| "logits/rejected": -0.7235937714576721, | |
| "logps/chosen": -188.78823852539062, | |
| "logps/rejected": -171.70339965820312, | |
| "loss": 0.8293, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.7909164428710938, | |
| "rewards/margins": 0.5921815037727356, | |
| "rewards/rejected": 1.198734998703003, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.27930763178599527, | |
| "grad_norm": 24.096284866333008, | |
| "learning_rate": 3.6359358540578095e-05, | |
| "logits/chosen": -0.30726513266563416, | |
| "logits/rejected": -0.8273059725761414, | |
| "logps/chosen": -199.73843383789062, | |
| "logps/rejected": -171.33981323242188, | |
| "loss": 0.7125, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.799203634262085, | |
| "rewards/margins": 0.8719380497932434, | |
| "rewards/rejected": 0.9272655248641968, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.2832415420928403, | |
| "grad_norm": 24.70514678955078, | |
| "learning_rate": 3.619970456233395e-05, | |
| "logits/chosen": -0.2640475332736969, | |
| "logits/rejected": -0.8327733278274536, | |
| "logps/chosen": -206.46035766601562, | |
| "logps/rejected": -165.4632110595703, | |
| "loss": 0.6377, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 2.429255723953247, | |
| "rewards/margins": 1.0934853553771973, | |
| "rewards/rejected": 1.3357700109481812, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2871754523996853, | |
| "grad_norm": 23.147350311279297, | |
| "learning_rate": 3.6036991104555974e-05, | |
| "logits/chosen": -0.008176600560545921, | |
| "logits/rejected": -0.34810671210289, | |
| "logps/chosen": -206.794921875, | |
| "logps/rejected": -186.4044189453125, | |
| "loss": 0.7023, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 2.2398250102996826, | |
| "rewards/margins": 0.7909523248672485, | |
| "rewards/rejected": 1.4488725662231445, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.2911093627065303, | |
| "grad_norm": 27.7244815826416, | |
| "learning_rate": 3.587124889734125e-05, | |
| "logits/chosen": -0.48228806257247925, | |
| "logits/rejected": -0.49171242117881775, | |
| "logps/chosen": -178.5115509033203, | |
| "logps/rejected": -172.22119140625, | |
| "loss": 0.9182, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.4874086380004883, | |
| "rewards/margins": 0.3070749342441559, | |
| "rewards/rejected": 1.1803337335586548, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2950432730133753, | |
| "grad_norm": 26.767961502075195, | |
| "learning_rate": 3.57025092427971e-05, | |
| "logits/chosen": -0.20643901824951172, | |
| "logits/rejected": -0.5273423790931702, | |
| "logps/chosen": -195.8766632080078, | |
| "logps/rejected": -173.40370178222656, | |
| "loss": 0.9323, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 2.199556827545166, | |
| "rewards/margins": 0.6209264993667603, | |
| "rewards/rejected": 1.5786300897598267, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.2989771833202203, | |
| "grad_norm": 18.721200942993164, | |
| "learning_rate": 3.5530804009129375e-05, | |
| "logits/chosen": -0.25294172763824463, | |
| "logits/rejected": -0.60113525390625, | |
| "logps/chosen": -201.85067749023438, | |
| "logps/rejected": -173.7031707763672, | |
| "loss": 0.7695, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 2.1115119457244873, | |
| "rewards/margins": 0.6209571361541748, | |
| "rewards/rejected": 1.4905551671981812, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3029110936270653, | |
| "grad_norm": 23.3524227142334, | |
| "learning_rate": 3.53561656246238e-05, | |
| "logits/chosen": -0.282802015542984, | |
| "logits/rejected": -0.5362482666969299, | |
| "logps/chosen": -202.23257446289062, | |
| "logps/rejected": -186.3799591064453, | |
| "loss": 0.681, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 2.020700454711914, | |
| "rewards/margins": 0.8641992807388306, | |
| "rewards/rejected": 1.156501293182373, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3068450039339103, | |
| "grad_norm": 19.006500244140625, | |
| "learning_rate": 3.5178627071521574e-05, | |
| "logits/chosen": -0.40107831358909607, | |
| "logits/rejected": -0.5502349138259888, | |
| "logps/chosen": -202.1154022216797, | |
| "logps/rejected": -185.59539794921875, | |
| "loss": 0.8749, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.338216781616211, | |
| "rewards/margins": 0.3761267364025116, | |
| "rewards/rejected": 0.9620901346206665, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3107789142407553, | |
| "grad_norm": 16.3823299407959, | |
| "learning_rate": 3.4998221879790324e-05, | |
| "logits/chosen": -0.31552472710609436, | |
| "logits/rejected": -0.7142918705940247, | |
| "logps/chosen": -197.4582977294922, | |
| "logps/rejected": -184.94174194335938, | |
| "loss": 0.901, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.363880455493927, | |
| "rewards/margins": 0.18122999370098114, | |
| "rewards/rejected": 0.18265047669410706, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3147128245476003, | |
| "grad_norm": 20.992412567138672, | |
| "learning_rate": 3.481498412079167e-05, | |
| "logits/chosen": -0.3988286852836609, | |
| "logits/rejected": -0.7025038003921509, | |
| "logps/chosen": -196.63803100585938, | |
| "logps/rejected": -178.83221435546875, | |
| "loss": 0.7963, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.26729291677474976, | |
| "rewards/margins": 0.3433241546154022, | |
| "rewards/rejected": -0.07603128254413605, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3147128245476003, | |
| "eval_logits/chosen": 1.2825068235397339, | |
| "eval_logits/rejected": 1.0665369033813477, | |
| "eval_logps/chosen": -207.7469940185547, | |
| "eval_logps/rejected": -180.9813232421875, | |
| "eval_loss": 0.8834741711616516, | |
| "eval_rewards/accuracies": 0.589062511920929, | |
| "eval_rewards/chosen": 0.08931777626276016, | |
| "eval_rewards/margins": 0.3287123441696167, | |
| "eval_rewards/rejected": -0.23939454555511475, | |
| "eval_runtime": 251.0719, | |
| "eval_samples_per_second": 2.549, | |
| "eval_steps_per_second": 0.159, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.31864673485444533, | |
| "grad_norm": 26.44162368774414, | |
| "learning_rate": 3.462894840084642e-05, | |
| "logits/chosen": -0.38823994994163513, | |
| "logits/rejected": -0.6459166407585144, | |
| "logps/chosen": -190.995361328125, | |
| "logps/rejected": -173.5061492919922, | |
| "loss": 1.0993, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.19754914939403534, | |
| "rewards/margins": -0.06637996435165405, | |
| "rewards/rejected": -0.1311691701412201, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 23.88953971862793, | |
| "learning_rate": 3.4440149854698904e-05, | |
| "logits/chosen": -0.38585788011550903, | |
| "logits/rejected": -0.7214224338531494, | |
| "logps/chosen": -205.794189453125, | |
| "logps/rejected": -175.1046600341797, | |
| "loss": 0.7178, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.010714983567595482, | |
| "rewards/margins": 0.5465508699417114, | |
| "rewards/rejected": -0.535835862159729, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.32651455546813535, | |
| "grad_norm": 22.045515060424805, | |
| "learning_rate": 3.424862413888134e-05, | |
| "logits/chosen": -0.30153313279151917, | |
| "logits/rejected": -0.49861517548561096, | |
| "logps/chosen": -193.66433715820312, | |
| "logps/rejected": -185.2410430908203, | |
| "loss": 0.7397, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7127935290336609, | |
| "rewards/margins": 0.6260470151901245, | |
| "rewards/rejected": 0.08674647659063339, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3304484657749803, | |
| "grad_norm": 27.552871704101562, | |
| "learning_rate": 3.4054407424979804e-05, | |
| "logits/chosen": -0.24517476558685303, | |
| "logits/rejected": -0.6450390815734863, | |
| "logps/chosen": -207.15200805664062, | |
| "logps/rejected": -176.4978485107422, | |
| "loss": 0.8156, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.253557562828064, | |
| "rewards/margins": 0.6033787131309509, | |
| "rewards/rejected": 0.6501787900924683, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.33438237608182536, | |
| "grad_norm": 17.781063079833984, | |
| "learning_rate": 3.3857536392802806e-05, | |
| "logits/chosen": -0.20200546085834503, | |
| "logits/rejected": -0.5061112642288208, | |
| "logps/chosen": -211.17922973632812, | |
| "logps/rejected": -192.77493286132812, | |
| "loss": 0.7883, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.2221243381500244, | |
| "rewards/margins": 0.416652113199234, | |
| "rewards/rejected": 0.805472195148468, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.33831628638867034, | |
| "grad_norm": 31.43796730041504, | |
| "learning_rate": 3.365804822345396e-05, | |
| "logits/chosen": -0.30611589550971985, | |
| "logits/rejected": -0.7340467572212219, | |
| "logps/chosen": -188.98428344726562, | |
| "logps/rejected": -158.50927734375, | |
| "loss": 0.7486, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 1.536189079284668, | |
| "rewards/margins": 0.589336633682251, | |
| "rewards/rejected": 0.9468523859977722, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3422501966955153, | |
| "grad_norm": 21.21323585510254, | |
| "learning_rate": 3.3455980592309925e-05, | |
| "logits/chosen": -0.445300817489624, | |
| "logits/rejected": -0.6968305110931396, | |
| "logps/chosen": -191.3743896484375, | |
| "logps/rejected": -173.6061248779297, | |
| "loss": 0.8092, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 2.1057379245758057, | |
| "rewards/margins": 0.6257677674293518, | |
| "rewards/rejected": 1.4799703359603882, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.34618410700236035, | |
| "grad_norm": 18.975866317749023, | |
| "learning_rate": 3.3251371661905066e-05, | |
| "logits/chosen": -0.05589728429913521, | |
| "logits/rejected": -0.5498273968696594, | |
| "logps/chosen": -201.5423583984375, | |
| "logps/rejected": -165.8180694580078, | |
| "loss": 0.6459, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 1.7923845052719116, | |
| "rewards/margins": 0.8942405581474304, | |
| "rewards/rejected": 0.8981439471244812, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.35011801730920533, | |
| "grad_norm": 25.26014518737793, | |
| "learning_rate": 3.3044260074724036e-05, | |
| "logits/chosen": -0.47154492139816284, | |
| "logits/rejected": -0.7864343523979187, | |
| "logps/chosen": -182.36541748046875, | |
| "logps/rejected": -167.81796264648438, | |
| "loss": 0.791, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.5660803318023682, | |
| "rewards/margins": 0.4582035541534424, | |
| "rewards/rejected": 1.1078766584396362, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.35405192761605037, | |
| "grad_norm": 20.178096771240234, | |
| "learning_rate": 3.283468494590378e-05, | |
| "logits/chosen": -0.5151780247688293, | |
| "logits/rejected": -0.8873647451400757, | |
| "logps/chosen": -186.1438751220703, | |
| "logps/rejected": -172.07485961914062, | |
| "loss": 0.7255, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.5227115154266357, | |
| "rewards/margins": 0.7071123123168945, | |
| "rewards/rejected": 0.815599262714386, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.35798583792289534, | |
| "grad_norm": 25.483036041259766, | |
| "learning_rate": 3.26226858558462e-05, | |
| "logits/chosen": -0.4173513948917389, | |
| "logits/rejected": -0.8169075846672058, | |
| "logps/chosen": -220.79647827148438, | |
| "logps/rejected": -183.74241638183594, | |
| "loss": 0.9569, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.048793911933899, | |
| "rewards/margins": 0.1925489604473114, | |
| "rewards/rejected": 0.8562449216842651, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.3619197482297404, | |
| "grad_norm": 24.004615783691406, | |
| "learning_rate": 3.240830284274301e-05, | |
| "logits/chosen": -0.5872361063957214, | |
| "logits/rejected": -0.9254258275032043, | |
| "logps/chosen": -199.5272674560547, | |
| "logps/rejected": -171.4918212890625, | |
| "loss": 0.8168, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.2763752043247223, | |
| "rewards/margins": 0.33582815527915955, | |
| "rewards/rejected": -0.05945297330617905, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.36585365853658536, | |
| "grad_norm": 23.19999885559082, | |
| "learning_rate": 3.219157639501416e-05, | |
| "logits/chosen": -0.2660069167613983, | |
| "logits/rejected": -0.4420255720615387, | |
| "logps/chosen": -187.30184936523438, | |
| "logps/rejected": -184.6049041748047, | |
| "loss": 0.871, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.41878828406333923, | |
| "rewards/margins": 0.461493581533432, | |
| "rewards/rejected": -0.04270533099770546, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.3697875688434304, | |
| "grad_norm": 26.874906539916992, | |
| "learning_rate": 3.1972547443661114e-05, | |
| "logits/chosen": -0.19552254676818848, | |
| "logits/rejected": -0.5767315030097961, | |
| "logps/chosen": -205.4563446044922, | |
| "logps/rejected": -186.74722290039062, | |
| "loss": 0.8292, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.7297551035881042, | |
| "rewards/margins": 0.3652102053165436, | |
| "rewards/rejected": 0.3645448684692383, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.37372147915027537, | |
| "grad_norm": 25.91361427307129, | |
| "learning_rate": 3.175125735453664e-05, | |
| "logits/chosen": -0.20148754119873047, | |
| "logits/rejected": -0.5481426119804382, | |
| "logps/chosen": -221.0868682861328, | |
| "logps/rejected": -197.3042449951172, | |
| "loss": 0.776, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.2072087526321411, | |
| "rewards/margins": 0.3937892019748688, | |
| "rewards/rejected": 0.8134196400642395, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.3776553894571204, | |
| "grad_norm": 20.111371994018555, | |
| "learning_rate": 3.152774792053247e-05, | |
| "logits/chosen": -0.3326271176338196, | |
| "logits/rejected": -0.5712831616401672, | |
| "logps/chosen": -203.29025268554688, | |
| "logps/rejected": -187.23904418945312, | |
| "loss": 0.691, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.6326103210449219, | |
| "rewards/margins": 0.8585765957832336, | |
| "rewards/rejected": 0.7740335464477539, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.3815892997639654, | |
| "grad_norm": 24.087158203125, | |
| "learning_rate": 3.1302061353686264e-05, | |
| "logits/chosen": -0.21493315696716309, | |
| "logits/rejected": -0.42921167612075806, | |
| "logps/chosen": -213.1800079345703, | |
| "logps/rejected": -195.15737915039062, | |
| "loss": 1.0061, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.751986026763916, | |
| "rewards/margins": 0.45154237747192383, | |
| "rewards/rejected": 2.3004441261291504, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.38552321007081036, | |
| "grad_norm": 37.55143356323242, | |
| "learning_rate": 3.107424027720941e-05, | |
| "logits/chosen": -0.5423256158828735, | |
| "logits/rejected": -1.0729395151138306, | |
| "logps/chosen": -193.28945922851562, | |
| "logps/rejected": -157.5052947998047, | |
| "loss": 0.7334, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 2.7436954975128174, | |
| "rewards/margins": 0.9733211398124695, | |
| "rewards/rejected": 1.7703745365142822, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.3894571203776554, | |
| "grad_norm": 24.617494583129883, | |
| "learning_rate": 3.0844327717437265e-05, | |
| "logits/chosen": -0.006182658486068249, | |
| "logits/rejected": -0.4748276174068451, | |
| "logps/chosen": -206.80899047851562, | |
| "logps/rejected": -172.13172912597656, | |
| "loss": 0.7678, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 2.979702949523926, | |
| "rewards/margins": 0.7632544636726379, | |
| "rewards/rejected": 2.2164485454559326, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.3933910306845004, | |
| "grad_norm": 17.355024337768555, | |
| "learning_rate": 3.0612367095703124e-05, | |
| "logits/chosen": -0.045905523002147675, | |
| "logits/rejected": -0.5579321384429932, | |
| "logps/chosen": -204.173583984375, | |
| "logps/rejected": -175.21615600585938, | |
| "loss": 0.6961, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 2.6483314037323, | |
| "rewards/margins": 0.7989780306816101, | |
| "rewards/rejected": 1.8493534326553345, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3933910306845004, | |
| "eval_logits/chosen": 1.3562368154525757, | |
| "eval_logits/rejected": 1.1292431354522705, | |
| "eval_logps/chosen": -202.31715393066406, | |
| "eval_logps/rejected": -176.03109741210938, | |
| "eval_loss": 0.8593710660934448, | |
| "eval_rewards/accuracies": 0.614062488079071, | |
| "eval_rewards/chosen": 2.261263370513916, | |
| "eval_rewards/margins": 0.520569920539856, | |
| "eval_rewards/rejected": 1.7406933307647705, | |
| "eval_runtime": 250.7548, | |
| "eval_samples_per_second": 2.552, | |
| "eval_steps_per_second": 0.16, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3973249409913454, | |
| "grad_norm": 21.848352432250977, | |
| "learning_rate": 3.0378402220137696e-05, | |
| "logits/chosen": 0.116512730717659, | |
| "logits/rejected": -0.3560892641544342, | |
| "logps/chosen": -202.67498779296875, | |
| "logps/rejected": -178.8372802734375, | |
| "loss": 0.7315, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 2.147618532180786, | |
| "rewards/margins": 0.7310143709182739, | |
| "rewards/rejected": 1.416603922843933, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.4012588512981904, | |
| "grad_norm": 27.533050537109375, | |
| "learning_rate": 3.0142477277395462e-05, | |
| "logits/chosen": 0.0075813233852386475, | |
| "logits/rejected": -0.6258127689361572, | |
| "logps/chosen": -209.9758758544922, | |
| "logps/rejected": -166.99766540527344, | |
| "loss": 0.6907, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.3229306936264038, | |
| "rewards/margins": 0.7913904190063477, | |
| "rewards/rejected": 0.5315402746200562, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4051927616050354, | |
| "grad_norm": 16.1999454498291, | |
| "learning_rate": 2.990463682430963e-05, | |
| "logits/chosen": -0.35700559616088867, | |
| "logits/rejected": -0.5182300209999084, | |
| "logps/chosen": -200.20985412597656, | |
| "logps/rejected": -179.12539672851562, | |
| "loss": 0.7671, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.9399553537368774, | |
| "rewards/margins": 0.48222360014915466, | |
| "rewards/rejected": 0.4577317237854004, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.4091266719118804, | |
| "grad_norm": 32.54240798950195, | |
| "learning_rate": 2.966492577947704e-05, | |
| "logits/chosen": -0.12386194616556168, | |
| "logits/rejected": -0.4458809494972229, | |
| "logps/chosen": -189.1624755859375, | |
| "logps/rejected": -168.4569091796875, | |
| "loss": 0.9158, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 1.4050761461257935, | |
| "rewards/margins": 0.6978380084037781, | |
| "rewards/rejected": 0.7072380781173706, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.41306058221872544, | |
| "grad_norm": 21.36524200439453, | |
| "learning_rate": 2.9423389414774918e-05, | |
| "logits/chosen": -0.18413856625556946, | |
| "logits/rejected": -0.5436551570892334, | |
| "logps/chosen": -206.7151336669922, | |
| "logps/rejected": -181.02978515625, | |
| "loss": 0.8962, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 1.010108232498169, | |
| "rewards/margins": 0.30978208780288696, | |
| "rewards/rejected": 0.700326144695282, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.4169944925255704, | |
| "grad_norm": 57.26128005981445, | |
| "learning_rate": 2.918007334681074e-05, | |
| "logits/chosen": -0.3969922661781311, | |
| "logits/rejected": -0.7256218194961548, | |
| "logps/chosen": -199.84579467773438, | |
| "logps/rejected": -165.75674438476562, | |
| "loss": 0.6673, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 2.075115442276001, | |
| "rewards/margins": 0.9931079745292664, | |
| "rewards/rejected": 1.0820075273513794, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4209284028324154, | |
| "grad_norm": 19.898866653442383, | |
| "learning_rate": 2.8935023528307124e-05, | |
| "logits/chosen": -0.1375611126422882, | |
| "logits/rejected": -0.4858548641204834, | |
| "logps/chosen": -201.7903289794922, | |
| "logps/rejected": -179.10194396972656, | |
| "loss": 0.7802, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 2.139580488204956, | |
| "rewards/margins": 0.8191660046577454, | |
| "rewards/rejected": 1.3204143047332764, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.42486231313926043, | |
| "grad_norm": 21.983154296875, | |
| "learning_rate": 2.868828623942317e-05, | |
| "logits/chosen": -0.16572633385658264, | |
| "logits/rejected": -0.5750501155853271, | |
| "logps/chosen": -197.57725524902344, | |
| "logps/rejected": -168.39517211914062, | |
| "loss": 0.5993, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 2.660771608352661, | |
| "rewards/margins": 1.5578687191009521, | |
| "rewards/rejected": 1.1029030084609985, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4287962234461054, | |
| "grad_norm": 25.60312843322754, | |
| "learning_rate": 2.8439908079013975e-05, | |
| "logits/chosen": -0.039677783846855164, | |
| "logits/rejected": -0.45520296692848206, | |
| "logps/chosen": -201.81983947753906, | |
| "logps/rejected": -175.03994750976562, | |
| "loss": 0.6675, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 2.3861243724823, | |
| "rewards/margins": 0.9951289892196655, | |
| "rewards/rejected": 1.3909955024719238, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.43273013375295044, | |
| "grad_norm": 18.380178451538086, | |
| "learning_rate": 2.8189935955829977e-05, | |
| "logits/chosen": -0.1264878809452057, | |
| "logits/rejected": -0.4807268977165222, | |
| "logps/chosen": -212.05575561523438, | |
| "logps/rejected": -190.66098022460938, | |
| "loss": 1.1466, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 1.8014596700668335, | |
| "rewards/margins": 0.04773964732885361, | |
| "rewards/rejected": 1.7537200450897217, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4366640440597954, | |
| "grad_norm": 23.785415649414062, | |
| "learning_rate": 2.7938417079657746e-05, | |
| "logits/chosen": -0.5862978100776672, | |
| "logits/rejected": -0.9251095056533813, | |
| "logps/chosen": -208.53695678710938, | |
| "logps/rejected": -180.97097778320312, | |
| "loss": 0.7145, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 2.073394775390625, | |
| "rewards/margins": 0.8073825836181641, | |
| "rewards/rejected": 1.2660123109817505, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.44059795436664045, | |
| "grad_norm": 24.911800384521484, | |
| "learning_rate": 2.7685398952403946e-05, | |
| "logits/chosen": -0.22417886555194855, | |
| "logits/rejected": -0.524285078048706, | |
| "logps/chosen": -207.4349365234375, | |
| "logps/rejected": -192.63758850097656, | |
| "loss": 0.9911, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 2.434218645095825, | |
| "rewards/margins": 0.4671500623226166, | |
| "rewards/rejected": 1.9670684337615967, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.44453186467348543, | |
| "grad_norm": 21.26664161682129, | |
| "learning_rate": 2.743092935912409e-05, | |
| "logits/chosen": -0.168512225151062, | |
| "logits/rejected": -0.4610822796821594, | |
| "logps/chosen": -207.57046508789062, | |
| "logps/rejected": -194.7887420654297, | |
| "loss": 0.9337, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 2.3753578662872314, | |
| "rewards/margins": 0.32185056805610657, | |
| "rewards/rejected": 2.0535073280334473, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.44846577498033047, | |
| "grad_norm": 23.40607452392578, | |
| "learning_rate": 2.7175056358997892e-05, | |
| "logits/chosen": -0.2319369614124298, | |
| "logits/rejected": -0.6572431325912476, | |
| "logps/chosen": -204.3990936279297, | |
| "logps/rejected": -180.0780487060547, | |
| "loss": 0.655, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 2.375079393386841, | |
| "rewards/margins": 0.9229291081428528, | |
| "rewards/rejected": 1.4521501064300537, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.45239968528717545, | |
| "grad_norm": 27.157663345336914, | |
| "learning_rate": 2.6917828276252747e-05, | |
| "logits/chosen": -0.1399884819984436, | |
| "logits/rejected": -0.46016645431518555, | |
| "logps/chosen": -218.91256713867188, | |
| "logps/rejected": -200.45068359375, | |
| "loss": 0.8026, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.7319839000701904, | |
| "rewards/margins": 0.786080539226532, | |
| "rewards/rejected": 0.9459033012390137, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.4563335955940205, | |
| "grad_norm": 20.441112518310547, | |
| "learning_rate": 2.6659293691037246e-05, | |
| "logits/chosen": -0.29733794927597046, | |
| "logits/rejected": -0.6675035953521729, | |
| "logps/chosen": -202.4237518310547, | |
| "logps/rejected": -176.8135223388672, | |
| "loss": 0.7622, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.0021812915802002, | |
| "rewards/margins": 0.6122254133224487, | |
| "rewards/rejected": 0.3899558186531067, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.46026750590086546, | |
| "grad_norm": 22.287860870361328, | |
| "learning_rate": 2.639950143024629e-05, | |
| "logits/chosen": -0.3219448924064636, | |
| "logits/rejected": -0.6755141019821167, | |
| "logps/chosen": -213.97811889648438, | |
| "logps/rejected": -184.30332946777344, | |
| "loss": 0.7878, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.9022668600082397, | |
| "rewards/margins": 0.5562304258346558, | |
| "rewards/rejected": 0.34603649377822876, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.46420141620771044, | |
| "grad_norm": 26.395917892456055, | |
| "learning_rate": 2.613850055829967e-05, | |
| "logits/chosen": -0.2554970383644104, | |
| "logits/rejected": -0.41837620735168457, | |
| "logps/chosen": -191.0037078857422, | |
| "logps/rejected": -185.1075897216797, | |
| "loss": 0.726, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.7684756517410278, | |
| "rewards/margins": 0.5839057564735413, | |
| "rewards/rejected": 0.18456986546516418, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.46813532651455547, | |
| "grad_norm": 25.363393783569336, | |
| "learning_rate": 2.587634036787571e-05, | |
| "logits/chosen": -0.37291693687438965, | |
| "logits/rejected": -0.9038352966308594, | |
| "logps/chosen": -202.77792358398438, | |
| "logps/rejected": -163.73086547851562, | |
| "loss": 0.7478, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 1.1803675889968872, | |
| "rewards/margins": 0.8836774826049805, | |
| "rewards/rejected": 0.2966901659965515, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.47206923682140045, | |
| "grad_norm": 27.41828727722168, | |
| "learning_rate": 2.5613070370601863e-05, | |
| "logits/chosen": -0.34387531876564026, | |
| "logits/rejected": -0.6873073577880859, | |
| "logps/chosen": -197.41146850585938, | |
| "logps/rejected": -171.6976318359375, | |
| "loss": 0.7599, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.3946095705032349, | |
| "rewards/margins": 0.7079658508300781, | |
| "rewards/rejected": 0.686643660068512, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.47206923682140045, | |
| "eval_logits/chosen": 1.3973848819732666, | |
| "eval_logits/rejected": 1.1919221878051758, | |
| "eval_logps/chosen": -203.7362518310547, | |
| "eval_logps/rejected": -177.5819854736328, | |
| "eval_loss": 0.8243551254272461, | |
| "eval_rewards/accuracies": 0.612500011920929, | |
| "eval_rewards/chosen": 1.693615198135376, | |
| "eval_rewards/margins": 0.5732673406600952, | |
| "eval_rewards/rejected": 1.1203477382659912, | |
| "eval_runtime": 251.0741, | |
| "eval_samples_per_second": 2.549, | |
| "eval_steps_per_second": 0.159, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4760031471282455, | |
| "grad_norm": 14.156159400939941, | |
| "learning_rate": 2.5348740287703942e-05, | |
| "logits/chosen": -0.28574198484420776, | |
| "logits/rejected": -0.7568629384040833, | |
| "logps/chosen": -192.17074584960938, | |
| "logps/rejected": -158.06005859375, | |
| "loss": 0.8055, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 1.7736876010894775, | |
| "rewards/margins": 0.6764134764671326, | |
| "rewards/rejected": 1.0972741842269897, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.47993705743509046, | |
| "grad_norm": 25.657800674438477, | |
| "learning_rate": 2.508340004061574e-05, | |
| "logits/chosen": -0.24258217215538025, | |
| "logits/rejected": -0.6088284254074097, | |
| "logps/chosen": -222.4447784423828, | |
| "logps/rejected": -190.37503051757812, | |
| "loss": 0.8069, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 2.226174831390381, | |
| "rewards/margins": 0.7204557657241821, | |
| "rewards/rejected": 1.5057189464569092, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.4838709677419355, | |
| "grad_norm": 28.052139282226562, | |
| "learning_rate": 2.4817099741550864e-05, | |
| "logits/chosen": -0.3004125952720642, | |
| "logits/rejected": -0.5441415309906006, | |
| "logps/chosen": -196.12718200683594, | |
| "logps/rejected": -192.7050018310547, | |
| "loss": 0.9248, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 2.162635326385498, | |
| "rewards/margins": 0.5665928721427917, | |
| "rewards/rejected": 1.5960423946380615, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.4878048780487805, | |
| "grad_norm": 23.494508743286133, | |
| "learning_rate": 2.4549889684038543e-05, | |
| "logits/chosen": -0.611892819404602, | |
| "logits/rejected": -0.8981421589851379, | |
| "logps/chosen": -190.71786499023438, | |
| "logps/rejected": -167.4793701171875, | |
| "loss": 0.7981, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 2.283334732055664, | |
| "rewards/margins": 0.6954591870307922, | |
| "rewards/rejected": 1.5878756046295166, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.4917387883556255, | |
| "grad_norm": 15.813286781311035, | |
| "learning_rate": 2.4281820333425172e-05, | |
| "logits/chosen": -0.19501128792762756, | |
| "logits/rejected": -0.6402938365936279, | |
| "logps/chosen": -194.93563842773438, | |
| "logps/rejected": -193.161376953125, | |
| "loss": 0.7841, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.6924636363983154, | |
| "rewards/margins": 0.6635173559188843, | |
| "rewards/rejected": 1.0289465188980103, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.4956726986624705, | |
| "grad_norm": 34.470821380615234, | |
| "learning_rate": 2.4012942317343404e-05, | |
| "logits/chosen": -0.10980845987796783, | |
| "logits/rejected": -0.5895272493362427, | |
| "logps/chosen": -207.9347381591797, | |
| "logps/rejected": -179.27322387695312, | |
| "loss": 0.8625, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.9872920513153076, | |
| "rewards/margins": 0.6902352571487427, | |
| "rewards/rejected": 1.2970569133758545, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.4996066089693155, | |
| "grad_norm": 21.860822677612305, | |
| "learning_rate": 2.374330641615064e-05, | |
| "logits/chosen": -0.47541099786758423, | |
| "logits/rejected": -0.6988755464553833, | |
| "logps/chosen": -181.64480590820312, | |
| "logps/rejected": -169.11415100097656, | |
| "loss": 1.0446, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.7799679636955261, | |
| "rewards/margins": -0.016168225556612015, | |
| "rewards/rejected": 0.796136200428009, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.5035405192761605, | |
| "grad_norm": 24.410993576049805, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "logits/chosen": -0.1209309846162796, | |
| "logits/rejected": -0.6129121780395508, | |
| "logps/chosen": -208.46646118164062, | |
| "logps/rejected": -173.28662109375, | |
| "loss": 0.7813, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.01343834400177, | |
| "rewards/margins": 0.7845452427864075, | |
| "rewards/rejected": 0.22889307141304016, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5074744295830055, | |
| "grad_norm": 16.415790557861328, | |
| "learning_rate": 2.3201964785915956e-05, | |
| "logits/chosen": -0.06124670431017876, | |
| "logits/rejected": -0.34292641282081604, | |
| "logps/chosen": -202.11863708496094, | |
| "logps/rejected": -185.01980590820312, | |
| "loss": 0.6251, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.7241964936256409, | |
| "rewards/margins": 0.7067764401435852, | |
| "rewards/rejected": 0.017420072108507156, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.5114083398898505, | |
| "grad_norm": 21.06545066833496, | |
| "learning_rate": 2.29303612947656e-05, | |
| "logits/chosen": -0.18204979598522186, | |
| "logits/rejected": -0.668160080909729, | |
| "logps/chosen": -217.71627807617188, | |
| "logps/rejected": -200.24777221679688, | |
| "loss": 0.6768, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.7872123718261719, | |
| "rewards/margins": 0.820330023765564, | |
| "rewards/rejected": -0.0331176295876503, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5153422501966956, | |
| "grad_norm": 20.972436904907227, | |
| "learning_rate": 2.265820437497871e-05, | |
| "logits/chosen": -0.4221726059913635, | |
| "logits/rejected": -0.7673285603523254, | |
| "logps/chosen": -193.30931091308594, | |
| "logps/rejected": -182.8597412109375, | |
| "loss": 0.5963, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.8721264600753784, | |
| "rewards/margins": 1.0373555421829224, | |
| "rewards/rejected": -0.1652289777994156, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.5192761605035405, | |
| "grad_norm": 25.041427612304688, | |
| "learning_rate": 2.2385545426167115e-05, | |
| "logits/chosen": -0.24139158427715302, | |
| "logits/rejected": -0.6950744390487671, | |
| "logps/chosen": -218.1499481201172, | |
| "logps/rejected": -176.1727752685547, | |
| "loss": 0.742, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 1.286029577255249, | |
| "rewards/margins": 0.7787684202194214, | |
| "rewards/rejected": 0.5072611570358276, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5232100708103855, | |
| "grad_norm": 24.626811981201172, | |
| "learning_rate": 2.2112435942755942e-05, | |
| "logits/chosen": -0.3278786540031433, | |
| "logits/rejected": -0.5997655987739563, | |
| "logps/chosen": -197.0006561279297, | |
| "logps/rejected": -180.835205078125, | |
| "loss": 0.9312, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 1.228043794631958, | |
| "rewards/margins": 0.33655455708503723, | |
| "rewards/rejected": 0.8914892077445984, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.5271439811172305, | |
| "grad_norm": 22.894763946533203, | |
| "learning_rate": 2.1838927504258357e-05, | |
| "logits/chosen": -0.251998633146286, | |
| "logits/rejected": -0.6962267160415649, | |
| "logps/chosen": -204.58847045898438, | |
| "logps/rejected": -177.03872680664062, | |
| "loss": 0.6696, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.4768571853637695, | |
| "rewards/margins": 0.7038243412971497, | |
| "rewards/rejected": 0.7730330228805542, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5310778914240756, | |
| "grad_norm": 18.778362274169922, | |
| "learning_rate": 2.156507176553429e-05, | |
| "logits/chosen": -0.1405915766954422, | |
| "logits/rejected": -0.6027745008468628, | |
| "logps/chosen": -204.18991088867188, | |
| "logps/rejected": -187.57095336914062, | |
| "loss": 0.5288, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 2.3066718578338623, | |
| "rewards/margins": 1.1950688362121582, | |
| "rewards/rejected": 1.111603021621704, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.5350118017309206, | |
| "grad_norm": 17.80322265625, | |
| "learning_rate": 2.129092044703485e-05, | |
| "logits/chosen": -0.15540140867233276, | |
| "logits/rejected": -0.4367486834526062, | |
| "logps/chosen": -197.77114868164062, | |
| "logps/rejected": -183.8221893310547, | |
| "loss": 0.8221, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 2.124110460281372, | |
| "rewards/margins": 0.5898113250732422, | |
| "rewards/rejected": 1.5342991352081299, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5389457120377655, | |
| "grad_norm": 20.969318389892578, | |
| "learning_rate": 2.1016525325034405e-05, | |
| "logits/chosen": -0.06777362525463104, | |
| "logits/rejected": -0.6237557530403137, | |
| "logps/chosen": -211.85391235351562, | |
| "logps/rejected": -179.53981018066406, | |
| "loss": 0.7552, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 2.0291833877563477, | |
| "rewards/margins": 0.694739818572998, | |
| "rewards/rejected": 1.33444344997406, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.5428796223446105, | |
| "grad_norm": 16.9830265045166, | |
| "learning_rate": 2.0741938221852104e-05, | |
| "logits/chosen": -0.3013477623462677, | |
| "logits/rejected": -0.6262307167053223, | |
| "logps/chosen": -188.5670166015625, | |
| "logps/rejected": -171.5750274658203, | |
| "loss": 0.7812, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 1.7201658487319946, | |
| "rewards/margins": 0.6786609888076782, | |
| "rewards/rejected": 1.0415048599243164, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5468135326514555, | |
| "grad_norm": 20.50164222717285, | |
| "learning_rate": 2.046721099606471e-05, | |
| "logits/chosen": -0.2895171046257019, | |
| "logits/rejected": -0.7959840893745422, | |
| "logps/chosen": -202.86526489257812, | |
| "logps/rejected": -168.116943359375, | |
| "loss": 0.6659, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 1.7198221683502197, | |
| "rewards/margins": 0.9006645083427429, | |
| "rewards/rejected": 0.8191574215888977, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.5507474429583006, | |
| "grad_norm": 24.72359848022461, | |
| "learning_rate": 2.0192395532712556e-05, | |
| "logits/chosen": -0.43050312995910645, | |
| "logits/rejected": -0.6164566278457642, | |
| "logps/chosen": -197.2266082763672, | |
| "logps/rejected": -185.25006103515625, | |
| "loss": 0.5887, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.4764689207077026, | |
| "rewards/margins": 0.8574800491333008, | |
| "rewards/rejected": 0.6189889311790466, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5507474429583006, | |
| "eval_logits/chosen": 1.277055025100708, | |
| "eval_logits/rejected": 1.059131145477295, | |
| "eval_logps/chosen": -205.26318359375, | |
| "eval_logps/rejected": -179.12974548339844, | |
| "eval_loss": 0.7933992743492126, | |
| "eval_rewards/accuracies": 0.6015625, | |
| "eval_rewards/chosen": 1.0828492641448975, | |
| "eval_rewards/margins": 0.5816105008125305, | |
| "eval_rewards/rejected": 0.5012389421463013, | |
| "eval_runtime": 250.5558, | |
| "eval_samples_per_second": 2.554, | |
| "eval_steps_per_second": 0.16, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5546813532651456, | |
| "grad_norm": 22.896121978759766, | |
| "learning_rate": 1.9917543733500532e-05, | |
| "logits/chosen": -0.005652183201164007, | |
| "logits/rejected": -0.21010038256645203, | |
| "logps/chosen": -193.43350219726562, | |
| "logps/rejected": -171.82823181152344, | |
| "loss": 0.8332, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.7293393015861511, | |
| "rewards/margins": 0.2607612609863281, | |
| "rewards/rejected": 0.4685780107975006, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.5586152635719905, | |
| "grad_norm": 19.870878219604492, | |
| "learning_rate": 1.9642707506995958e-05, | |
| "logits/chosen": -0.15198977291584015, | |
| "logits/rejected": -0.5298476815223694, | |
| "logps/chosen": -211.7928466796875, | |
| "logps/rejected": -179.28244018554688, | |
| "loss": 0.5933, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.4447071552276611, | |
| "rewards/margins": 1.0455639362335205, | |
| "rewards/rejected": 0.39914312958717346, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.5625491738788355, | |
| "grad_norm": 23.537195205688477, | |
| "learning_rate": 1.9367938758825056e-05, | |
| "logits/chosen": -0.1395382583141327, | |
| "logits/rejected": -0.6614929437637329, | |
| "logps/chosen": -209.0814208984375, | |
| "logps/rejected": -173.17478942871094, | |
| "loss": 0.6686, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.0658318996429443, | |
| "rewards/margins": 0.7761461734771729, | |
| "rewards/rejected": 0.2896856367588043, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.5664830841856806, | |
| "grad_norm": 22.774639129638672, | |
| "learning_rate": 1.90932893818701e-05, | |
| "logits/chosen": -0.09201023727655411, | |
| "logits/rejected": -0.3696451187133789, | |
| "logps/chosen": -216.9640655517578, | |
| "logps/rejected": -189.9328155517578, | |
| "loss": 0.9453, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 1.0307306051254272, | |
| "rewards/margins": 0.025239372625947, | |
| "rewards/rejected": 1.0054912567138672, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5704169944925256, | |
| "grad_norm": 20.04373550415039, | |
| "learning_rate": 1.8818811246468875e-05, | |
| "logits/chosen": -0.21691718697547913, | |
| "logits/rejected": -0.6853747367858887, | |
| "logps/chosen": -197.96017456054688, | |
| "logps/rejected": -165.71035766601562, | |
| "loss": 0.7018, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 1.1925420761108398, | |
| "rewards/margins": 0.6861337423324585, | |
| "rewards/rejected": 0.5064083337783813, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.5743509047993706, | |
| "grad_norm": 21.558950424194336, | |
| "learning_rate": 1.8544556190618464e-05, | |
| "logits/chosen": 0.01636449061334133, | |
| "logits/rejected": -0.25090569257736206, | |
| "logps/chosen": -206.47286987304688, | |
| "logps/rejected": -188.0323486328125, | |
| "loss": 0.7587, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 1.4961506128311157, | |
| "rewards/margins": 0.7738775014877319, | |
| "rewards/rejected": 0.722273051738739, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.5782848151062155, | |
| "grad_norm": 25.04125213623047, | |
| "learning_rate": 1.8270576010185092e-05, | |
| "logits/chosen": -0.3876519501209259, | |
| "logits/rejected": -0.7680070996284485, | |
| "logps/chosen": -203.02491760253906, | |
| "logps/rejected": -169.53672790527344, | |
| "loss": 0.7619, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.2711286544799805, | |
| "rewards/margins": 0.6064616441726685, | |
| "rewards/rejected": 0.6646669507026672, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.5822187254130606, | |
| "grad_norm": 19.16695213317871, | |
| "learning_rate": 1.799692244912195e-05, | |
| "logits/chosen": -0.1374569684267044, | |
| "logits/rejected": -0.48932093381881714, | |
| "logps/chosen": -223.61813354492188, | |
| "logps/rejected": -199.4268341064453, | |
| "loss": 0.6542, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.8322197198867798, | |
| "rewards/margins": 1.0043301582336426, | |
| "rewards/rejected": 0.8278897404670715, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5861526357199056, | |
| "grad_norm": 22.487895965576172, | |
| "learning_rate": 1.7723647189696843e-05, | |
| "logits/chosen": -0.11542626470327377, | |
| "logits/rejected": -0.3874351382255554, | |
| "logps/chosen": -205.8312225341797, | |
| "logps/rejected": -180.7718505859375, | |
| "loss": 0.8241, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.5367833375930786, | |
| "rewards/margins": 0.5126625895500183, | |
| "rewards/rejected": 1.0241206884384155, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.5900865460267506, | |
| "grad_norm": 16.18448257446289, | |
| "learning_rate": 1.7450801842731445e-05, | |
| "logits/chosen": -0.023925015702843666, | |
| "logits/rejected": -0.4763374328613281, | |
| "logps/chosen": -214.9409942626953, | |
| "logps/rejected": -185.83316040039062, | |
| "loss": 0.623, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.6252682209014893, | |
| "rewards/margins": 1.056118369102478, | |
| "rewards/rejected": 0.5691498517990112, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5940204563335956, | |
| "grad_norm": 28.91587257385254, | |
| "learning_rate": 1.7178437937854065e-05, | |
| "logits/chosen": -0.1288028210401535, | |
| "logits/rejected": -0.6050774455070496, | |
| "logps/chosen": -207.6809539794922, | |
| "logps/rejected": -183.9476776123047, | |
| "loss": 0.7653, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 1.3927785158157349, | |
| "rewards/margins": 0.8300971984863281, | |
| "rewards/rejected": 0.5626811981201172, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.5979543666404405, | |
| "grad_norm": 18.849876403808594, | |
| "learning_rate": 1.6906606913767778e-05, | |
| "logits/chosen": -0.2586264908313751, | |
| "logits/rejected": -0.6829530596733093, | |
| "logps/chosen": -200.44479370117188, | |
| "logps/rejected": -167.65695190429688, | |
| "loss": 0.7795, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.2213537693023682, | |
| "rewards/margins": 0.4475148618221283, | |
| "rewards/rejected": 0.7738388180732727, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6018882769472856, | |
| "grad_norm": 25.64194679260254, | |
| "learning_rate": 1.663536010853567e-05, | |
| "logits/chosen": -0.28778618574142456, | |
| "logits/rejected": -0.8489567637443542, | |
| "logps/chosen": -210.17996215820312, | |
| "logps/rejected": -162.97006225585938, | |
| "loss": 0.7059, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 1.7119789123535156, | |
| "rewards/margins": 1.09046471118927, | |
| "rewards/rejected": 0.6215142011642456, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.6058221872541306, | |
| "grad_norm": 14.11388874053955, | |
| "learning_rate": 1.6364748749885137e-05, | |
| "logits/chosen": -0.12569937109947205, | |
| "logits/rejected": -0.47604307532310486, | |
| "logps/chosen": -211.08639526367188, | |
| "logps/rejected": -186.719970703125, | |
| "loss": 0.7375, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.7021243572235107, | |
| "rewards/margins": 0.5951262712478638, | |
| "rewards/rejected": 1.106998085975647, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6097560975609756, | |
| "grad_norm": 20.6273193359375, | |
| "learning_rate": 1.6094823945533e-05, | |
| "logits/chosen": 0.008131015114486217, | |
| "logits/rejected": -0.5134211778640747, | |
| "logps/chosen": -211.6152801513672, | |
| "logps/rejected": -173.49942016601562, | |
| "loss": 0.8714, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.374456763267517, | |
| "rewards/margins": 0.7731674909591675, | |
| "rewards/rejected": 0.6012891530990601, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.6136900078678206, | |
| "grad_norm": 17.514083862304688, | |
| "learning_rate": 1.58256366735333e-05, | |
| "logits/chosen": -0.0192732997238636, | |
| "logits/rejected": -0.37163615226745605, | |
| "logps/chosen": -192.7857208251953, | |
| "logps/rejected": -184.69699096679688, | |
| "loss": 0.5552, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.496361255645752, | |
| "rewards/margins": 1.1403007507324219, | |
| "rewards/rejected": 0.35606056451797485, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6176239181746657, | |
| "grad_norm": 12.284652709960938, | |
| "learning_rate": 1.555723777264957e-05, | |
| "logits/chosen": -0.4136200547218323, | |
| "logits/rejected": -0.7878348231315613, | |
| "logps/chosen": -188.60914611816406, | |
| "logps/rejected": -162.43040466308594, | |
| "loss": 0.7347, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 1.2863472700119019, | |
| "rewards/margins": 0.7062928676605225, | |
| "rewards/rejected": 0.580054521560669, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.6215578284815106, | |
| "grad_norm": 26.990276336669922, | |
| "learning_rate": 1.52896779327534e-05, | |
| "logits/chosen": -0.062058307230472565, | |
| "logits/rejected": -0.4678223133087158, | |
| "logps/chosen": -197.87442016601562, | |
| "logps/rejected": -181.07705688476562, | |
| "loss": 0.7406, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 1.5467268228530884, | |
| "rewards/margins": 0.6265745162963867, | |
| "rewards/rejected": 0.9201523661613464, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6254917387883556, | |
| "grad_norm": 26.78017807006836, | |
| "learning_rate": 1.5023007685251153e-05, | |
| "logits/chosen": -0.37522611021995544, | |
| "logits/rejected": -0.6946735382080078, | |
| "logps/chosen": -197.1485137939453, | |
| "logps/rejected": -177.7647705078125, | |
| "loss": 0.8528, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.4908908605575562, | |
| "rewards/margins": 0.8086551427841187, | |
| "rewards/rejected": 0.682235598564148, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.6294256490952006, | |
| "grad_norm": 22.932809829711914, | |
| "learning_rate": 1.4757277393540602e-05, | |
| "logits/chosen": -0.07537481188774109, | |
| "logits/rejected": -0.4230824112892151, | |
| "logps/chosen": -195.38076782226562, | |
| "logps/rejected": -173.30007934570312, | |
| "loss": 0.6356, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 1.180471420288086, | |
| "rewards/margins": 0.8355466723442078, | |
| "rewards/rejected": 0.3449247181415558, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6294256490952006, | |
| "eval_logits/chosen": 1.2686213254928589, | |
| "eval_logits/rejected": 1.0565236806869507, | |
| "eval_logps/chosen": -205.04721069335938, | |
| "eval_logps/rejected": -179.17825317382812, | |
| "eval_loss": 0.8022397756576538, | |
| "eval_rewards/accuracies": 0.628125011920929, | |
| "eval_rewards/chosen": 1.1692273616790771, | |
| "eval_rewards/margins": 0.6873818039894104, | |
| "eval_rewards/rejected": 0.4818454682826996, | |
| "eval_runtime": 251.1114, | |
| "eval_samples_per_second": 2.549, | |
| "eval_steps_per_second": 0.159, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6333595594020456, | |
| "grad_norm": 27.06431007385254, | |
| "learning_rate": 1.4492537243499253e-05, | |
| "logits/chosen": -0.08504633605480194, | |
| "logits/rejected": -0.37000229954719543, | |
| "logps/chosen": -221.24990844726562, | |
| "logps/rejected": -189.64706420898438, | |
| "loss": 0.8898, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.172984004020691, | |
| "rewards/margins": 0.574467658996582, | |
| "rewards/rejected": 0.5985165238380432, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.6372934697088907, | |
| "grad_norm": 20.225252151489258, | |
| "learning_rate": 1.4228837234006273e-05, | |
| "logits/chosen": -0.09346888214349747, | |
| "logits/rejected": -0.43980512022972107, | |
| "logps/chosen": -197.91098022460938, | |
| "logps/rejected": -192.83395385742188, | |
| "loss": 0.6881, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.971549391746521, | |
| "rewards/margins": 0.6740128397941589, | |
| "rewards/rejected": 0.29753655195236206, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6412273800157356, | |
| "grad_norm": 28.533573150634766, | |
| "learning_rate": 1.3966227167499668e-05, | |
| "logits/chosen": 0.0283651240170002, | |
| "logits/rejected": -0.32705655694007874, | |
| "logps/chosen": -229.0713348388672, | |
| "logps/rejected": -199.23040771484375, | |
| "loss": 0.8796, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.845023512840271, | |
| "rewards/margins": 0.3233472406864166, | |
| "rewards/rejected": 0.5216764211654663, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 18.14284896850586, | |
| "learning_rate": 1.3704756640570575e-05, | |
| "logits/chosen": -0.26658809185028076, | |
| "logits/rejected": -0.6564346551895142, | |
| "logps/chosen": -194.92213439941406, | |
| "logps/rejected": -176.8538818359375, | |
| "loss": 0.5813, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.0112560987472534, | |
| "rewards/margins": 0.9789012670516968, | |
| "rewards/rejected": 0.03235483542084694, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6490952006294256, | |
| "grad_norm": 25.14547348022461, | |
| "learning_rate": 1.3444475034596464e-05, | |
| "logits/chosen": 0.08274314552545547, | |
| "logits/rejected": -0.47508057951927185, | |
| "logps/chosen": -189.21493530273438, | |
| "logps/rejected": -163.44387817382812, | |
| "loss": 0.7053, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 1.0577847957611084, | |
| "rewards/margins": 0.9744871258735657, | |
| "rewards/rejected": 0.08329786360263824, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.6530291109362707, | |
| "grad_norm": 23.569334030151367, | |
| "learning_rate": 1.3185431506414945e-05, | |
| "logits/chosen": -0.2618711590766907, | |
| "logits/rejected": -0.6126792430877686, | |
| "logps/chosen": -197.77426147460938, | |
| "logps/rejected": -172.6019744873047, | |
| "loss": 0.6912, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.0107938051223755, | |
| "rewards/margins": 0.7313990592956543, | |
| "rewards/rejected": 0.2793947160243988, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6569630212431157, | |
| "grad_norm": 14.946830749511719, | |
| "learning_rate": 1.292767497904001e-05, | |
| "logits/chosen": 0.09540750831365585, | |
| "logits/rejected": -0.42838025093078613, | |
| "logps/chosen": -214.9687957763672, | |
| "logps/rejected": -183.12184143066406, | |
| "loss": 0.6963, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.7165493965148926, | |
| "rewards/margins": 0.6403694152832031, | |
| "rewards/rejected": 0.07618004828691483, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.6608969315499607, | |
| "grad_norm": 17.816455841064453, | |
| "learning_rate": 1.2671254132422393e-05, | |
| "logits/chosen": -0.34346064925193787, | |
| "logits/rejected": -0.628952145576477, | |
| "logps/chosen": -192.6859588623047, | |
| "logps/rejected": -179.92324829101562, | |
| "loss": 0.5884, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.1312153339385986, | |
| "rewards/margins": 0.8917140960693359, | |
| "rewards/rejected": 0.2395012080669403, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.6648308418568056, | |
| "grad_norm": 25.672109603881836, | |
| "learning_rate": 1.2416217394255906e-05, | |
| "logits/chosen": -0.2732202410697937, | |
| "logits/rejected": -0.5638844966888428, | |
| "logps/chosen": -207.419921875, | |
| "logps/rejected": -190.7101287841797, | |
| "loss": 0.8709, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.9203730821609497, | |
| "rewards/margins": 0.5037940740585327, | |
| "rewards/rejected": 0.4165789484977722, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.6687647521636507, | |
| "grad_norm": 24.7791805267334, | |
| "learning_rate": 1.2162612930831357e-05, | |
| "logits/chosen": -0.021216195076704025, | |
| "logits/rejected": -0.4218730032444, | |
| "logps/chosen": -211.8462677001953, | |
| "logps/rejected": -183.61422729492188, | |
| "loss": 0.6667, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.6997087001800537, | |
| "rewards/margins": 0.6386712193489075, | |
| "rewards/rejected": 0.06103747338056564, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6726986624704957, | |
| "grad_norm": 17.186410903930664, | |
| "learning_rate": 1.1910488637939826e-05, | |
| "logits/chosen": -0.19025997817516327, | |
| "logits/rejected": -0.36145836114883423, | |
| "logps/chosen": -191.56838989257812, | |
| "logps/rejected": -188.1061553955078, | |
| "loss": 0.7408, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.7108181715011597, | |
| "rewards/margins": 0.6494898200035095, | |
| "rewards/rejected": 0.061328280717134476, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.6766325727773407, | |
| "grad_norm": 18.120513916015625, | |
| "learning_rate": 1.16598921318271e-05, | |
| "logits/chosen": -0.21306462585926056, | |
| "logits/rejected": -0.5272704362869263, | |
| "logps/chosen": -198.18896484375, | |
| "logps/rejected": -180.7800750732422, | |
| "loss": 0.6794, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.7278915643692017, | |
| "rewards/margins": 0.8657029271125793, | |
| "rewards/rejected": -0.13781137764453888, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.6805664830841857, | |
| "grad_norm": 20.887371063232422, | |
| "learning_rate": 1.1410870740200841e-05, | |
| "logits/chosen": -0.3551502823829651, | |
| "logits/rejected": -0.7362738847732544, | |
| "logps/chosen": -193.05130004882812, | |
| "logps/rejected": -174.00880432128906, | |
| "loss": 0.7762, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.8120868802070618, | |
| "rewards/margins": 0.8929181098937988, | |
| "rewards/rejected": -0.08083119988441467, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.6845003933910306, | |
| "grad_norm": 24.89902687072754, | |
| "learning_rate": 1.1163471493292268e-05, | |
| "logits/chosen": -0.009045323356986046, | |
| "logits/rejected": -0.34948888421058655, | |
| "logps/chosen": -220.1766815185547, | |
| "logps/rejected": -188.1958770751953, | |
| "loss": 0.9159, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 1.148590326309204, | |
| "rewards/margins": 0.5903881192207336, | |
| "rewards/rejected": 0.5582022666931152, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6884343036978757, | |
| "grad_norm": 20.535686492919922, | |
| "learning_rate": 1.0917741114974007e-05, | |
| "logits/chosen": -0.3664267361164093, | |
| "logits/rejected": -0.8371219635009766, | |
| "logps/chosen": -200.46461486816406, | |
| "logps/rejected": -163.69137573242188, | |
| "loss": 0.7615, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.84300297498703, | |
| "rewards/margins": 0.8857911825180054, | |
| "rewards/rejected": -0.04278818517923355, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.6923682140047207, | |
| "grad_norm": 27.465072631835938, | |
| "learning_rate": 1.0673726013935828e-05, | |
| "logits/chosen": -0.25900501012802124, | |
| "logits/rejected": -0.7321861982345581, | |
| "logps/chosen": -211.7742919921875, | |
| "logps/rejected": -171.26019287109375, | |
| "loss": 0.8606, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.3226665258407593, | |
| "rewards/margins": 0.6031711101531982, | |
| "rewards/rejected": 0.7194954752922058, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6963021243115657, | |
| "grad_norm": 24.614782333374023, | |
| "learning_rate": 1.0431472274919864e-05, | |
| "logits/chosen": 0.13692393898963928, | |
| "logits/rejected": -0.47194236516952515, | |
| "logps/chosen": -215.8333282470703, | |
| "logps/rejected": -175.57333374023438, | |
| "loss": 0.7391, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.1747673749923706, | |
| "rewards/margins": 0.6339668035507202, | |
| "rewards/rejected": 0.5408006906509399, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.7002360346184107, | |
| "grad_norm": 24.77642250061035, | |
| "learning_rate": 1.019102565001707e-05, | |
| "logits/chosen": -0.4733211100101471, | |
| "logits/rejected": -0.7766138315200806, | |
| "logps/chosen": -199.7375946044922, | |
| "logps/rejected": -169.17922973632812, | |
| "loss": 0.7491, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.6666286587715149, | |
| "rewards/margins": 0.6510328054428101, | |
| "rewards/rejected": 0.015595817938446999, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7041699449252558, | |
| "grad_norm": 14.021129608154297, | |
| "learning_rate": 9.95243155002646e-06, | |
| "logits/chosen": -0.18163976073265076, | |
| "logits/rejected": -0.785339891910553, | |
| "logps/chosen": -198.67337036132812, | |
| "logps/rejected": -167.85330200195312, | |
| "loss": 0.5539, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.2840198278427124, | |
| "rewards/margins": 1.1790852546691895, | |
| "rewards/rejected": 0.10493452847003937, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.7081038552321007, | |
| "grad_norm": 27.48203468322754, | |
| "learning_rate": 9.7157350358788e-06, | |
| "logits/chosen": -0.5614138245582581, | |
| "logits/rejected": -0.7469819188117981, | |
| "logps/chosen": -204.990478515625, | |
| "logps/rejected": -186.52420043945312, | |
| "loss": 0.8492, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.8738373517990112, | |
| "rewards/margins": 0.5437658429145813, | |
| "rewards/rejected": 0.3300715386867523, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7081038552321007, | |
| "eval_logits/chosen": 1.3100758790969849, | |
| "eval_logits/rejected": 1.0955623388290405, | |
| "eval_logps/chosen": -205.67874145507812, | |
| "eval_logps/rejected": -179.78384399414062, | |
| "eval_loss": 0.75962895154953, | |
| "eval_rewards/accuracies": 0.628125011920929, | |
| "eval_rewards/chosen": 0.9166278839111328, | |
| "eval_rewards/margins": 0.6770327091217041, | |
| "eval_rewards/rejected": 0.23959510028362274, | |
| "eval_runtime": 250.4549, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 0.16, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7120377655389457, | |
| "grad_norm": 16.553695678710938, | |
| "learning_rate": 9.480980810126412e-06, | |
| "logits/chosen": -0.16247500479221344, | |
| "logits/rejected": -0.5344902276992798, | |
| "logps/chosen": -183.8999786376953, | |
| "logps/rejected": -169.76052856445312, | |
| "loss": 0.7105, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.8913670778274536, | |
| "rewards/margins": 0.8846918344497681, | |
| "rewards/rejected": 0.006675288081169128, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.7159716758457907, | |
| "grad_norm": 28.125730514526367, | |
| "learning_rate": 9.248213208500629e-06, | |
| "logits/chosen": -0.03472292423248291, | |
| "logits/rejected": -0.5100794434547424, | |
| "logps/chosen": -223.66665649414062, | |
| "logps/rejected": -193.7150421142578, | |
| "loss": 0.6846, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 1.2616242170333862, | |
| "rewards/margins": 0.8945581316947937, | |
| "rewards/rejected": 0.36706608533859253, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7199055861526357, | |
| "grad_norm": 20.356327056884766, | |
| "learning_rate": 9.017476191538556e-06, | |
| "logits/chosen": 0.017731910571455956, | |
| "logits/rejected": -0.5705962181091309, | |
| "logps/chosen": -204.4373779296875, | |
| "logps/rejected": -166.6776885986328, | |
| "loss": 0.6665, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.9296242594718933, | |
| "rewards/margins": 0.7391864061355591, | |
| "rewards/rejected": 0.19043782353401184, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.7238394964594808, | |
| "grad_norm": 13.494754791259766, | |
| "learning_rate": 8.78881333628063e-06, | |
| "logits/chosen": -0.2843412160873413, | |
| "logits/rejected": -0.5937708616256714, | |
| "logps/chosen": -184.79656982421875, | |
| "logps/rejected": -171.83572387695312, | |
| "loss": 0.6447, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.7906737327575684, | |
| "rewards/margins": 0.704569935798645, | |
| "rewards/rejected": 0.08610378205776215, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7277734067663257, | |
| "grad_norm": 17.869462966918945, | |
| "learning_rate": 8.562267828040714e-06, | |
| "logits/chosen": -0.5717681050300598, | |
| "logits/rejected": -0.875778079032898, | |
| "logps/chosen": -192.8079071044922, | |
| "logps/rejected": -175.69139099121094, | |
| "loss": 0.7083, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.3932053744792938, | |
| "rewards/margins": 0.5822644233703613, | |
| "rewards/rejected": -0.1890590488910675, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.7317073170731707, | |
| "grad_norm": 18.533781051635742, | |
| "learning_rate": 8.337882452250058e-06, | |
| "logits/chosen": -0.34088677167892456, | |
| "logits/rejected": -0.6980454325675964, | |
| "logps/chosen": -210.4141082763672, | |
| "logps/rejected": -196.22714233398438, | |
| "loss": 0.8137, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.802699863910675, | |
| "rewards/margins": 0.6218786835670471, | |
| "rewards/rejected": 0.1808212548494339, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7356412273800157, | |
| "grad_norm": 20.780742645263672, | |
| "learning_rate": 8.115699586376865e-06, | |
| "logits/chosen": -0.5672314167022705, | |
| "logits/rejected": -0.855423629283905, | |
| "logps/chosen": -175.87796020507812, | |
| "logps/rejected": -159.25332641601562, | |
| "loss": 0.8287, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.42161092162132263, | |
| "rewards/margins": 0.5529305934906006, | |
| "rewards/rejected": -0.13131961226463318, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.7395751376868608, | |
| "grad_norm": 17.732147216796875, | |
| "learning_rate": 7.895761191922861e-06, | |
| "logits/chosen": -0.09196851402521133, | |
| "logits/rejected": -0.5184351801872253, | |
| "logps/chosen": -218.05563354492188, | |
| "logps/rejected": -186.40463256835938, | |
| "loss": 0.5937, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.9370241165161133, | |
| "rewards/margins": 0.9796267747879028, | |
| "rewards/rejected": -0.04260266199707985, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.7435090479937058, | |
| "grad_norm": 22.87380599975586, | |
| "learning_rate": 7.678108806498442e-06, | |
| "logits/chosen": -0.350538432598114, | |
| "logits/rejected": -0.6340444684028625, | |
| "logps/chosen": -202.5361785888672, | |
| "logps/rejected": -178.80300903320312, | |
| "loss": 0.7974, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.3468557596206665, | |
| "rewards/margins": 0.8867467641830444, | |
| "rewards/rejected": 0.46010905504226685, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.7474429583005507, | |
| "grad_norm": 23.245569229125977, | |
| "learning_rate": 7.462783535977842e-06, | |
| "logits/chosen": -0.19628608226776123, | |
| "logits/rejected": -0.6630419492721558, | |
| "logps/chosen": -206.7510223388672, | |
| "logps/rejected": -168.88804626464844, | |
| "loss": 0.8134, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.6633275747299194, | |
| "rewards/margins": 0.356858491897583, | |
| "rewards/rejected": 0.3064691126346588, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7513768686073957, | |
| "grad_norm": 25.233848571777344, | |
| "learning_rate": 7.249826046735928e-06, | |
| "logits/chosen": -0.2122970074415207, | |
| "logits/rejected": -0.6993425488471985, | |
| "logps/chosen": -222.96298217773438, | |
| "logps/rejected": -183.36917114257812, | |
| "loss": 0.8162, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 1.5078728199005127, | |
| "rewards/margins": 0.9431095123291016, | |
| "rewards/rejected": 0.5647634267807007, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.7553107789142408, | |
| "grad_norm": 23.390539169311523, | |
| "learning_rate": 7.039276557967895e-06, | |
| "logits/chosen": -0.17560990154743195, | |
| "logits/rejected": -0.5167850852012634, | |
| "logps/chosen": -210.7447967529297, | |
| "logps/rejected": -185.9241180419922, | |
| "loss": 0.7632, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.9249579310417175, | |
| "rewards/margins": 0.6043773889541626, | |
| "rewards/rejected": 0.32058054208755493, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.7592446892210858, | |
| "grad_norm": 21.24551773071289, | |
| "learning_rate": 6.831174834093477e-06, | |
| "logits/chosen": -0.20156054198741913, | |
| "logits/rejected": -0.6375221610069275, | |
| "logps/chosen": -203.7855987548828, | |
| "logps/rejected": -169.75369262695312, | |
| "loss": 0.6731, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.9981558918952942, | |
| "rewards/margins": 0.8284910321235657, | |
| "rewards/rejected": 0.16966481506824493, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.7631785995279308, | |
| "grad_norm": 21.75351333618164, | |
| "learning_rate": 6.625560177247023e-06, | |
| "logits/chosen": -0.15350133180618286, | |
| "logits/rejected": -0.4431697726249695, | |
| "logps/chosen": -185.0416717529297, | |
| "logps/rejected": -163.0596160888672, | |
| "loss": 0.8273, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 1.2868894338607788, | |
| "rewards/margins": 0.6607948541641235, | |
| "rewards/rejected": 0.6260946989059448, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.7671125098347757, | |
| "grad_norm": 21.936416625976562, | |
| "learning_rate": 6.422471419854899e-06, | |
| "logits/chosen": -0.4399512708187103, | |
| "logits/rejected": -0.6968399882316589, | |
| "logps/chosen": -190.70333862304688, | |
| "logps/rejected": -173.41281127929688, | |
| "loss": 0.6566, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 1.1954008340835571, | |
| "rewards/margins": 0.9580972790718079, | |
| "rewards/rejected": 0.23730352520942688, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.7710464201416207, | |
| "grad_norm": 21.626394271850586, | |
| "learning_rate": 6.221946917301563e-06, | |
| "logits/chosen": -0.012787247076630592, | |
| "logits/rejected": -0.2723299562931061, | |
| "logps/chosen": -212.55581665039062, | |
| "logps/rejected": -191.74188232421875, | |
| "loss": 0.7072, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.4022397994995117, | |
| "rewards/margins": 0.6362533569335938, | |
| "rewards/rejected": 0.765986442565918, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.7749803304484658, | |
| "grad_norm": 14.10751724243164, | |
| "learning_rate": 6.024024540685802e-06, | |
| "logits/chosen": -0.38578885793685913, | |
| "logits/rejected": -0.8282561302185059, | |
| "logps/chosen": -190.1944122314453, | |
| "logps/rejected": -164.04934692382812, | |
| "loss": 0.5289, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 1.3787574768066406, | |
| "rewards/margins": 0.9574896097183228, | |
| "rewards/rejected": 0.4212678372859955, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.7789142407553108, | |
| "grad_norm": 16.705045700073242, | |
| "learning_rate": 5.828741669668339e-06, | |
| "logits/chosen": -0.5691582560539246, | |
| "logits/rejected": -0.9283183813095093, | |
| "logps/chosen": -188.06460571289062, | |
| "logps/rejected": -168.5616455078125, | |
| "loss": 0.6278, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 1.050636887550354, | |
| "rewards/margins": 0.8141539692878723, | |
| "rewards/rejected": 0.23648293316364288, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.7828481510621558, | |
| "grad_norm": 18.042753219604492, | |
| "learning_rate": 5.636135185412342e-06, | |
| "logits/chosen": 0.26870280504226685, | |
| "logits/rejected": -0.28138986229896545, | |
| "logps/chosen": -204.53746032714844, | |
| "logps/rejected": -172.84890747070312, | |
| "loss": 0.636, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 1.0944331884384155, | |
| "rewards/margins": 0.9463173151016235, | |
| "rewards/rejected": 0.14811599254608154, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.7867820613690008, | |
| "grad_norm": 18.450891494750977, | |
| "learning_rate": 5.446241463618027e-06, | |
| "logits/chosen": -0.13302814960479736, | |
| "logits/rejected": -0.6271988153457642, | |
| "logps/chosen": -222.73452758789062, | |
| "logps/rejected": -177.82113647460938, | |
| "loss": 0.6449, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.3603211641311646, | |
| "rewards/margins": 0.9086629748344421, | |
| "rewards/rejected": 0.4516581594944, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7867820613690008, | |
| "eval_logits/chosen": 1.2821887731552124, | |
| "eval_logits/rejected": 1.063094973564148, | |
| "eval_logps/chosen": -204.9945526123047, | |
| "eval_logps/rejected": -179.13351440429688, | |
| "eval_loss": 0.7391278147697449, | |
| "eval_rewards/accuracies": 0.637499988079071, | |
| "eval_rewards/chosen": 1.190301775932312, | |
| "eval_rewards/margins": 0.6905705332756042, | |
| "eval_rewards/rejected": 0.49973124265670776, | |
| "eval_runtime": 249.286, | |
| "eval_samples_per_second": 2.567, | |
| "eval_steps_per_second": 0.16, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7907159716758458, | |
| "grad_norm": 22.258323669433594, | |
| "learning_rate": 5.2590963676527255e-06, | |
| "logits/chosen": -0.10886897891759872, | |
| "logits/rejected": -0.31225308775901794, | |
| "logps/chosen": -210.9266357421875, | |
| "logps/rejected": -194.21224975585938, | |
| "loss": 0.6442, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 1.1041626930236816, | |
| "rewards/margins": 0.7658023834228516, | |
| "rewards/rejected": 0.33836036920547485, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.7946498819826908, | |
| "grad_norm": 20.957698822021484, | |
| "learning_rate": 5.074735241777733e-06, | |
| "logits/chosen": -0.2896588146686554, | |
| "logits/rejected": -0.6819514036178589, | |
| "logps/chosen": -207.27120971679688, | |
| "logps/rejected": -179.46530151367188, | |
| "loss": 0.7299, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.9105139970779419, | |
| "rewards/margins": 0.4902356266975403, | |
| "rewards/rejected": 0.42027825117111206, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.7985837922895358, | |
| "grad_norm": 17.720888137817383, | |
| "learning_rate": 4.893192904473183e-06, | |
| "logits/chosen": -0.3812113106250763, | |
| "logits/rejected": -0.9288043975830078, | |
| "logps/chosen": -194.83285522460938, | |
| "logps/rejected": -157.9988250732422, | |
| "loss": 0.8882, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.6736253499984741, | |
| "rewards/margins": 0.4428628087043762, | |
| "rewards/rejected": 0.2307625561952591, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.8025177025963808, | |
| "grad_norm": 17.797956466674805, | |
| "learning_rate": 4.714503641862225e-06, | |
| "logits/chosen": -0.3434005677700043, | |
| "logits/rejected": -0.9068504571914673, | |
| "logps/chosen": -207.9601287841797, | |
| "logps/rejected": -173.23727416992188, | |
| "loss": 0.5452, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.209609031677246, | |
| "rewards/margins": 1.1478995084762573, | |
| "rewards/rejected": 0.06170947104692459, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.8064516129032258, | |
| "grad_norm": 16.82356834411621, | |
| "learning_rate": 4.538701201235713e-06, | |
| "logits/chosen": -0.22616150975227356, | |
| "logits/rejected": -0.6889921426773071, | |
| "logps/chosen": -192.7413787841797, | |
| "logps/rejected": -163.8821258544922, | |
| "loss": 0.6258, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 1.1558220386505127, | |
| "rewards/margins": 0.9748827219009399, | |
| "rewards/rejected": 0.18093928694725037, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.8103855232100708, | |
| "grad_norm": 20.977785110473633, | |
| "learning_rate": 4.365818784678737e-06, | |
| "logits/chosen": -0.03155326843261719, | |
| "logits/rejected": -0.41564369201660156, | |
| "logps/chosen": -208.5790252685547, | |
| "logps/rejected": -175.08924865722656, | |
| "loss": 0.7402, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 1.0165379047393799, | |
| "rewards/margins": 0.8640995025634766, | |
| "rewards/rejected": 0.15243832767009735, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.8143194335169158, | |
| "grad_norm": 32.46459197998047, | |
| "learning_rate": 4.195889042800021e-06, | |
| "logits/chosen": -0.17535440623760223, | |
| "logits/rejected": -0.4597929120063782, | |
| "logps/chosen": -199.391845703125, | |
| "logps/rejected": -179.8048553466797, | |
| "loss": 0.7109, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 1.0733716487884521, | |
| "rewards/margins": 0.6777491569519043, | |
| "rewards/rejected": 0.3956224024295807, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.8182533438237608, | |
| "grad_norm": 16.850854873657227, | |
| "learning_rate": 4.028944068565552e-06, | |
| "logits/chosen": -0.5630426406860352, | |
| "logits/rejected": -0.8235554695129395, | |
| "logps/chosen": -179.50892639160156, | |
| "logps/rejected": -167.5303955078125, | |
| "loss": 0.6135, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.1456284523010254, | |
| "rewards/margins": 0.971184253692627, | |
| "rewards/rejected": 0.17444416880607605, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8221872541306058, | |
| "grad_norm": 27.865501403808594, | |
| "learning_rate": 3.865015391237481e-06, | |
| "logits/chosen": -0.4081362783908844, | |
| "logits/rejected": -0.7251068353652954, | |
| "logps/chosen": -206.79672241210938, | |
| "logps/rejected": -183.33836364746094, | |
| "loss": 1.0297, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.3163181245326996, | |
| "rewards/margins": 0.19530019164085388, | |
| "rewards/rejected": 0.1210179552435875, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.8261211644374509, | |
| "grad_norm": 20.659204483032227, | |
| "learning_rate": 3.7041339704195147e-06, | |
| "logits/chosen": -0.1373288780450821, | |
| "logits/rejected": -0.5759511590003967, | |
| "logps/chosen": -203.7024383544922, | |
| "logps/rejected": -173.4119110107422, | |
| "loss": 0.8511, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.6036697626113892, | |
| "rewards/margins": 0.6436252593994141, | |
| "rewards/rejected": -0.0399555079638958, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8300550747442959, | |
| "grad_norm": 24.464290618896484, | |
| "learning_rate": 3.54633019020985e-06, | |
| "logits/chosen": -0.43429121375083923, | |
| "logits/rejected": -0.5019878149032593, | |
| "logps/chosen": -198.01043701171875, | |
| "logps/rejected": -187.71475219726562, | |
| "loss": 0.9095, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": 0.7602452039718628, | |
| "rewards/margins": 0.32361331582069397, | |
| "rewards/rejected": 0.436631977558136, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.8339889850511408, | |
| "grad_norm": 23.853845596313477, | |
| "learning_rate": 3.3916338534628613e-06, | |
| "logits/chosen": -0.11409912258386612, | |
| "logits/rejected": -0.40145158767700195, | |
| "logps/chosen": -210.5789794921875, | |
| "logps/rejected": -191.952392578125, | |
| "loss": 0.8087, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.8562873005867004, | |
| "rewards/margins": 0.44316864013671875, | |
| "rewards/rejected": 0.41311874985694885, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.8379228953579858, | |
| "grad_norm": 30.748018264770508, | |
| "learning_rate": 3.2400741761605102e-06, | |
| "logits/chosen": -0.21002812683582306, | |
| "logits/rejected": -0.5244247913360596, | |
| "logps/chosen": -195.76846313476562, | |
| "logps/rejected": -186.1009063720703, | |
| "loss": 0.7539, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.2099284678697586, | |
| "rewards/margins": 0.5657997131347656, | |
| "rewards/rejected": -0.35587114095687866, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.8418568056648308, | |
| "grad_norm": 20.117977142333984, | |
| "learning_rate": 3.0916797818946055e-06, | |
| "logits/chosen": -0.3361959457397461, | |
| "logits/rejected": -0.5092878937721252, | |
| "logps/chosen": -195.49034118652344, | |
| "logps/rejected": -186.6125030517578, | |
| "loss": 0.6731, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.40774106979370117, | |
| "rewards/margins": 0.6332494616508484, | |
| "rewards/rejected": -0.22550833225250244, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.8457907159716759, | |
| "grad_norm": 20.158287048339844, | |
| "learning_rate": 2.946478696460957e-06, | |
| "logits/chosen": -0.54124516248703, | |
| "logits/rejected": -0.8028567433357239, | |
| "logps/chosen": -192.8404998779297, | |
| "logps/rejected": -176.3671112060547, | |
| "loss": 0.6739, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.6862714886665344, | |
| "rewards/margins": 0.8415164947509766, | |
| "rewards/rejected": -0.15524490177631378, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.8497246262785209, | |
| "grad_norm": 17.690227508544922, | |
| "learning_rate": 2.8044983425664064e-06, | |
| "logits/chosen": -0.28623318672180176, | |
| "logits/rejected": -0.46538248658180237, | |
| "logps/chosen": -198.58580017089844, | |
| "logps/rejected": -183.89404296875, | |
| "loss": 0.7617, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.34613844752311707, | |
| "rewards/margins": 0.7363790273666382, | |
| "rewards/rejected": -0.3902406394481659, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.8536585365853658, | |
| "grad_norm": 26.19361114501953, | |
| "learning_rate": 2.6657655346497802e-06, | |
| "logits/chosen": -0.38322368264198303, | |
| "logits/rejected": -0.6140966415405273, | |
| "logps/chosen": -187.73886108398438, | |
| "logps/rejected": -176.44613647460938, | |
| "loss": 0.7017, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.20231589674949646, | |
| "rewards/margins": 0.5345470905303955, | |
| "rewards/rejected": -0.3322312533855438, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.8575924468922108, | |
| "grad_norm": 25.393978118896484, | |
| "learning_rate": 2.530306473817696e-06, | |
| "logits/chosen": -0.25616469979286194, | |
| "logits/rejected": -0.7887105941772461, | |
| "logps/chosen": -188.26930236816406, | |
| "logps/rejected": -164.24330139160156, | |
| "loss": 0.7071, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.39394354820251465, | |
| "rewards/margins": 0.7766583561897278, | |
| "rewards/rejected": -0.3827148377895355, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.8615263571990559, | |
| "grad_norm": 24.819536209106445, | |
| "learning_rate": 2.398146742896237e-06, | |
| "logits/chosen": -0.2826923131942749, | |
| "logits/rejected": -0.8120182156562805, | |
| "logps/chosen": -175.20675659179688, | |
| "logps/rejected": -142.05787658691406, | |
| "loss": 0.6828, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.2799919545650482, | |
| "rewards/margins": 0.5129075646400452, | |
| "rewards/rejected": -0.23291556537151337, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.8654602675059009, | |
| "grad_norm": 15.595181465148926, | |
| "learning_rate": 2.269311301599344e-06, | |
| "logits/chosen": -0.25191831588745117, | |
| "logits/rejected": -0.5568451881408691, | |
| "logps/chosen": -191.5902862548828, | |
| "logps/rejected": -171.51724243164062, | |
| "loss": 0.7207, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.6834554672241211, | |
| "rewards/margins": 0.6807907223701477, | |
| "rewards/rejected": 0.002664747880771756, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8654602675059009, | |
| "eval_logits/chosen": 1.2452117204666138, | |
| "eval_logits/rejected": 1.0254409313201904, | |
| "eval_logps/chosen": -206.24465942382812, | |
| "eval_logps/rejected": -180.29928588867188, | |
| "eval_loss": 0.737010657787323, | |
| "eval_rewards/accuracies": 0.6390625238418579, | |
| "eval_rewards/chosen": 0.6902583837509155, | |
| "eval_rewards/margins": 0.6568381786346436, | |
| "eval_rewards/rejected": 0.03342027962207794, | |
| "eval_runtime": 250.2397, | |
| "eval_samples_per_second": 2.558, | |
| "eval_steps_per_second": 0.16, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8693941778127459, | |
| "grad_norm": 26.02237319946289, | |
| "learning_rate": 2.143824481814947e-06, | |
| "logits/chosen": -0.5257942080497742, | |
| "logits/rejected": -0.6871519088745117, | |
| "logps/chosen": -183.809814453125, | |
| "logps/rejected": -175.24085998535156, | |
| "loss": 0.9311, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.2924796938896179, | |
| "rewards/margins": 0.16722363233566284, | |
| "rewards/rejected": 0.1252560317516327, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.8733280881195908, | |
| "grad_norm": 15.61811637878418, | |
| "learning_rate": 2.02170998300963e-06, | |
| "logits/chosen": -0.49455881118774414, | |
| "logits/rejected": -0.8319517970085144, | |
| "logps/chosen": -195.03097534179688, | |
| "logps/rejected": -172.79736328125, | |
| "loss": 0.7795, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.6870075464248657, | |
| "rewards/margins": 0.6901518702507019, | |
| "rewards/rejected": -0.0031443715561181307, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.8772619984264359, | |
| "grad_norm": 19.150976181030273, | |
| "learning_rate": 1.9029908677527409e-06, | |
| "logits/chosen": -0.27715402841567993, | |
| "logits/rejected": -0.5119122266769409, | |
| "logps/chosen": -190.5804901123047, | |
| "logps/rejected": -170.06210327148438, | |
| "loss": 0.7519, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.7375507950782776, | |
| "rewards/margins": 0.4395205080509186, | |
| "rewards/rejected": 0.2980303168296814, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.8811959087332809, | |
| "grad_norm": 20.794376373291016, | |
| "learning_rate": 1.78768955736079e-06, | |
| "logits/chosen": -0.47320452332496643, | |
| "logits/rejected": -0.5798605680465698, | |
| "logps/chosen": -186.94198608398438, | |
| "logps/rejected": -193.33441162109375, | |
| "loss": 0.837, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.7762082815170288, | |
| "rewards/margins": 0.520244836807251, | |
| "rewards/rejected": 0.25596338510513306, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.8851298190401259, | |
| "grad_norm": 23.86858558654785, | |
| "learning_rate": 1.6758278276629659e-06, | |
| "logits/chosen": -0.23235611617565155, | |
| "logits/rejected": -0.6398773789405823, | |
| "logps/chosen": -212.32974243164062, | |
| "logps/rejected": -180.15377807617188, | |
| "loss": 0.7666, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.5019949078559875, | |
| "rewards/margins": 0.5942084193229675, | |
| "rewards/rejected": -0.09221355617046356, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.8890637293469709, | |
| "grad_norm": 23.230112075805664, | |
| "learning_rate": 1.5674268048885277e-06, | |
| "logits/chosen": -0.23109452426433563, | |
| "logits/rejected": -0.593435525894165, | |
| "logps/chosen": -202.44102478027344, | |
| "logps/rejected": -179.34506225585938, | |
| "loss": 0.8569, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.6775950193405151, | |
| "rewards/margins": 0.16297699511051178, | |
| "rewards/rejected": 0.5146180391311646, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.8929976396538158, | |
| "grad_norm": 22.992965698242188, | |
| "learning_rate": 1.4625069616769215e-06, | |
| "logits/chosen": -0.3653388023376465, | |
| "logits/rejected": -0.7005013227462769, | |
| "logps/chosen": -199.02430725097656, | |
| "logps/rejected": -180.14822387695312, | |
| "loss": 0.7866, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.8748451471328735, | |
| "rewards/margins": 0.41802239418029785, | |
| "rewards/rejected": 0.45682281255722046, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.8969315499606609, | |
| "grad_norm": 13.673257827758789, | |
| "learning_rate": 1.3610881132113107e-06, | |
| "logits/chosen": -0.1476505994796753, | |
| "logits/rejected": -0.5143739581108093, | |
| "logps/chosen": -194.36900329589844, | |
| "logps/rejected": -174.2411651611328, | |
| "loss": 0.5425, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.6795452237129211, | |
| "rewards/margins": 1.169248104095459, | |
| "rewards/rejected": -0.4897027909755707, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.9008654602675059, | |
| "grad_norm": 28.896047592163086, | |
| "learning_rate": 1.2631894134762579e-06, | |
| "logits/chosen": -0.26152271032333374, | |
| "logits/rejected": -0.6222779750823975, | |
| "logps/chosen": -215.73617553710938, | |
| "logps/rejected": -192.5132293701172, | |
| "loss": 0.7018, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.757038950920105, | |
| "rewards/margins": 0.659284234046936, | |
| "rewards/rejected": 0.09775470197200775, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.9047993705743509, | |
| "grad_norm": 27.06733512878418, | |
| "learning_rate": 1.168829351640326e-06, | |
| "logits/chosen": 0.007974958047270775, | |
| "logits/rejected": -0.3661263585090637, | |
| "logps/chosen": -207.5963592529297, | |
| "logps/rejected": -186.1758575439453, | |
| "loss": 0.7401, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.5918732285499573, | |
| "rewards/margins": 0.6995636820793152, | |
| "rewards/rejected": -0.1076904684305191, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9087332808811959, | |
| "grad_norm": 32.632076263427734, | |
| "learning_rate": 1.078025748564191e-06, | |
| "logits/chosen": -0.11743499338626862, | |
| "logits/rejected": -0.08067210018634796, | |
| "logps/chosen": -205.4510498046875, | |
| "logps/rejected": -210.25064086914062, | |
| "loss": 0.9985, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.5534142851829529, | |
| "rewards/margins": -0.008044779300689697, | |
| "rewards/rejected": 0.5614590644836426, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.912667191188041, | |
| "grad_norm": 28.71379280090332, | |
| "learning_rate": 9.907957534349921e-07, | |
| "logits/chosen": -0.004957390017807484, | |
| "logits/rejected": -0.4194992482662201, | |
| "logps/chosen": -205.16690063476562, | |
| "logps/rejected": -183.73019409179688, | |
| "loss": 0.7593, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.7620848417282104, | |
| "rewards/margins": 0.7305535078048706, | |
| "rewards/rejected": 0.03153129667043686, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.9166011014948859, | |
| "grad_norm": 16.41864776611328, | |
| "learning_rate": 9.071558405275427e-07, | |
| "logits/chosen": -0.1401735246181488, | |
| "logits/rejected": -0.5280479192733765, | |
| "logps/chosen": -191.03903198242188, | |
| "logps/rejected": -167.55050659179688, | |
| "loss": 0.6841, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.8857030868530273, | |
| "rewards/margins": 0.8941364288330078, | |
| "rewards/rejected": -0.008433381095528603, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.9205350118017309, | |
| "grad_norm": 23.13312339782715, | |
| "learning_rate": 8.271218060929919e-07, | |
| "logits/chosen": 0.09978056699037552, | |
| "logits/rejected": -0.262615829706192, | |
| "logps/chosen": -214.7543487548828, | |
| "logps/rejected": -196.4529266357422, | |
| "loss": 0.6888, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.8993024826049805, | |
| "rewards/margins": 0.7381424903869629, | |
| "rewards/rejected": 0.1611599177122116, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.9244689221085759, | |
| "grad_norm": 18.770814895629883, | |
| "learning_rate": 7.507087653755318e-07, | |
| "logits/chosen": -0.07238979637622833, | |
| "logits/rejected": -0.5419026613235474, | |
| "logps/chosen": -209.29220581054688, | |
| "logps/rejected": -173.19552612304688, | |
| "loss": 0.7126, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.8157307505607605, | |
| "rewards/margins": 0.6763442158699036, | |
| "rewards/rejected": 0.13938648998737335, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.9284028324154209, | |
| "grad_norm": 25.199766159057617, | |
| "learning_rate": 6.779311497577401e-07, | |
| "logits/chosen": -0.2254069745540619, | |
| "logits/rejected": -0.4307781159877777, | |
| "logps/chosen": -196.3717803955078, | |
| "logps/rejected": -181.4769287109375, | |
| "loss": 0.7135, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.47276610136032104, | |
| "rewards/margins": 0.584132969379425, | |
| "rewards/rejected": -0.11136679351329803, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.932336742722266, | |
| "grad_norm": 18.558359146118164, | |
| "learning_rate": 6.088027040350674e-07, | |
| "logits/chosen": -0.4991328716278076, | |
| "logits/rejected": -0.9202349781990051, | |
| "logps/chosen": -180.4629669189453, | |
| "logps/rejected": -145.46389770507812, | |
| "loss": 0.7549, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.6303231120109558, | |
| "rewards/margins": 0.5688610672950745, | |
| "rewards/rejected": 0.06146197393536568, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.9362706530291109, | |
| "grad_norm": 29.797143936157227, | |
| "learning_rate": 5.433364838199828e-07, | |
| "logits/chosen": -0.06067700311541557, | |
| "logits/rejected": -0.5211640000343323, | |
| "logps/chosen": -220.21395874023438, | |
| "logps/rejected": -198.33328247070312, | |
| "loss": 0.856, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.7763031125068665, | |
| "rewards/margins": 0.5708464980125427, | |
| "rewards/rejected": 0.20545658469200134, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.9402045633359559, | |
| "grad_norm": 30.696592330932617, | |
| "learning_rate": 4.815448530762923e-07, | |
| "logits/chosen": -0.014176195487380028, | |
| "logits/rejected": -0.40428978204727173, | |
| "logps/chosen": -191.35374450683594, | |
| "logps/rejected": -164.81369018554688, | |
| "loss": 0.8069, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.6798957586288452, | |
| "rewards/margins": 0.4587259888648987, | |
| "rewards/rejected": 0.22116975486278534, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.9441384736428009, | |
| "grad_norm": 19.7831974029541, | |
| "learning_rate": 4.2343948178408036e-07, | |
| "logits/chosen": -0.5212418437004089, | |
| "logits/rejected": -0.7373995780944824, | |
| "logps/chosen": -217.49252319335938, | |
| "logps/rejected": -194.1028289794922, | |
| "loss": 0.7382, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.7597753405570984, | |
| "rewards/margins": 0.6988024711608887, | |
| "rewards/rejected": 0.0609729178249836, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9441384736428009, | |
| "eval_logits/chosen": 1.2533395290374756, | |
| "eval_logits/rejected": 1.0344760417938232, | |
| "eval_logps/chosen": -206.20431518554688, | |
| "eval_logps/rejected": -180.233154296875, | |
| "eval_loss": 0.7342613339424133, | |
| "eval_rewards/accuracies": 0.6421874761581421, | |
| "eval_rewards/chosen": 0.7063881158828735, | |
| "eval_rewards/margins": 0.6465141773223877, | |
| "eval_rewards/rejected": 0.05987401679158211, | |
| "eval_runtime": 251.029, | |
| "eval_samples_per_second": 2.55, | |
| "eval_steps_per_second": 0.159, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.948072383949646, | |
| "grad_norm": 15.179329872131348, | |
| "learning_rate": 3.6903134373571515e-07, | |
| "logits/chosen": 0.04393022134900093, | |
| "logits/rejected": -0.3573564887046814, | |
| "logps/chosen": -224.6765899658203, | |
| "logps/rejected": -196.52413940429688, | |
| "loss": 0.9152, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.5112829804420471, | |
| "rewards/margins": 0.32752710580825806, | |
| "rewards/rejected": 0.18375587463378906, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.952006294256491, | |
| "grad_norm": 26.592243194580078, | |
| "learning_rate": 3.1833071446333297e-07, | |
| "logits/chosen": -0.09226761013269424, | |
| "logits/rejected": -0.5401033163070679, | |
| "logps/chosen": -207.3560028076172, | |
| "logps/rejected": -179.0234832763672, | |
| "loss": 0.7916, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.7182900905609131, | |
| "rewards/margins": 0.3775801360607147, | |
| "rewards/rejected": 0.34070995450019836, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.955940204563336, | |
| "grad_norm": 17.654870986938477, | |
| "learning_rate": 2.713471692982017e-07, | |
| "logits/chosen": -0.10396593809127808, | |
| "logits/rejected": -0.5638905167579651, | |
| "logps/chosen": -221.9078369140625, | |
| "logps/rejected": -190.99435424804688, | |
| "loss": 0.6688, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 1.0176299810409546, | |
| "rewards/margins": 0.687278151512146, | |
| "rewards/rejected": 0.33035191893577576, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.9598741148701809, | |
| "grad_norm": 28.622743606567383, | |
| "learning_rate": 2.2808958156231853e-07, | |
| "logits/chosen": -0.3307565152645111, | |
| "logits/rejected": -0.6672178506851196, | |
| "logps/chosen": -205.37741088867188, | |
| "logps/rejected": -174.81289672851562, | |
| "loss": 0.8025, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.7133423089981079, | |
| "rewards/margins": 0.618229866027832, | |
| "rewards/rejected": 0.09511242806911469, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.963808025177026, | |
| "grad_norm": 20.119291305541992, | |
| "learning_rate": 1.8856612089259486e-07, | |
| "logits/chosen": -0.11120174080133438, | |
| "logits/rejected": -0.47498565912246704, | |
| "logps/chosen": -222.4508056640625, | |
| "logps/rejected": -193.77328491210938, | |
| "loss": 0.7055, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.6713041663169861, | |
| "rewards/margins": 0.6344063878059387, | |
| "rewards/rejected": 0.036897819489240646, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 23.203277587890625, | |
| "learning_rate": 1.5278425169794163e-07, | |
| "logits/chosen": -0.31409674882888794, | |
| "logits/rejected": -0.7144479155540466, | |
| "logps/chosen": -209.11270141601562, | |
| "logps/rejected": -189.7498779296875, | |
| "loss": 0.7145, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.8724400401115417, | |
| "rewards/margins": 0.7660844922065735, | |
| "rewards/rejected": 0.10635566711425781, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.971675845790716, | |
| "grad_norm": 17.31842803955078, | |
| "learning_rate": 1.2075073174952378e-07, | |
| "logits/chosen": -0.29665330052375793, | |
| "logits/rejected": -0.7147069573402405, | |
| "logps/chosen": -180.279541015625, | |
| "logps/rejected": -157.10206604003906, | |
| "loss": 0.7291, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.6150587797164917, | |
| "rewards/margins": 0.5622903108596802, | |
| "rewards/rejected": 0.052768461406230927, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.975609756097561, | |
| "grad_norm": 22.84340476989746, | |
| "learning_rate": 9.247161090451207e-08, | |
| "logits/chosen": -0.1943736970424652, | |
| "logits/rejected": -0.5411825180053711, | |
| "logps/chosen": -210.853759765625, | |
| "logps/rejected": -190.47271728515625, | |
| "loss": 0.7743, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.7618738412857056, | |
| "rewards/margins": 0.6717807054519653, | |
| "rewards/rejected": 0.09009318053722382, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.9795436664044059, | |
| "grad_norm": 18.27547836303711, | |
| "learning_rate": 6.795222996347495e-08, | |
| "logits/chosen": -0.16004569828510284, | |
| "logits/rejected": -0.5410211682319641, | |
| "logps/chosen": -222.00424194335938, | |
| "logps/rejected": -190.4541015625, | |
| "loss": 0.6775, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.933633029460907, | |
| "rewards/margins": 0.7526041269302368, | |
| "rewards/rejected": 0.18102897703647614, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.983477576711251, | |
| "grad_norm": 33.558162689208984, | |
| "learning_rate": 4.7197219661743176e-08, | |
| "logits/chosen": -0.308822900056839, | |
| "logits/rejected": -0.6432265043258667, | |
| "logps/chosen": -195.4747314453125, | |
| "logps/rejected": -166.3302459716797, | |
| "loss": 0.7417, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.7237275838851929, | |
| "rewards/margins": 0.6491376161575317, | |
| "rewards/rejected": 0.07458990067243576, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.987411487018096, | |
| "grad_norm": 16.65594482421875, | |
| "learning_rate": 3.021049979482715e-08, | |
| "logits/chosen": -0.6164910793304443, | |
| "logits/rejected": -0.9252687692642212, | |
| "logps/chosen": -188.57847595214844, | |
| "logps/rejected": -167.00997924804688, | |
| "loss": 0.823, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.2205405980348587, | |
| "rewards/margins": 0.4816465377807617, | |
| "rewards/rejected": -0.2611059546470642, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.991345397324941, | |
| "grad_norm": 21.41892433166504, | |
| "learning_rate": 1.6995278478133536e-08, | |
| "logits/chosen": -0.32881253957748413, | |
| "logits/rejected": -0.6159898042678833, | |
| "logps/chosen": -203.57620239257812, | |
| "logps/rejected": -189.83755493164062, | |
| "loss": 0.6967, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.2210671603679657, | |
| "rewards/margins": 0.6746386289596558, | |
| "rewards/rejected": -0.45357149839401245, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.995279307631786, | |
| "grad_norm": 25.134788513183594, | |
| "learning_rate": 7.554051541074359e-09, | |
| "logits/chosen": -0.724000096321106, | |
| "logits/rejected": -0.9451119303703308, | |
| "logps/chosen": -185.7470703125, | |
| "logps/rejected": -176.31640625, | |
| "loss": 0.7505, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.5909692049026489, | |
| "rewards/margins": 0.6639338731765747, | |
| "rewards/rejected": -0.07296473532915115, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.999213217938631, | |
| "grad_norm": 20.82672119140625, | |
| "learning_rate": 1.888602055710731e-09, | |
| "logits/chosen": -0.15106138586997986, | |
| "logits/rejected": -0.6675028800964355, | |
| "logps/chosen": -208.9075927734375, | |
| "logps/rejected": -172.70423889160156, | |
| "loss": 0.7455, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7576273679733276, | |
| "rewards/margins": 0.660781741142273, | |
| "rewards/rejected": 0.09684564918279648, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1271, | |
| "total_flos": 0.0, | |
| "train_loss": 0.7529747673697963, | |
| "train_runtime": 21716.0887, | |
| "train_samples_per_second": 0.936, | |
| "train_steps_per_second": 0.059 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1271, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |