| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9958330532965927, | |
| "eval_steps": 100, | |
| "global_step": 464, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004301364339001277, | |
| "grad_norm": 64.5691909790039, | |
| "learning_rate": 1.0638297872340425e-08, | |
| "logits/chosen": -1.1015625, | |
| "logits/rejected": -1.109375, | |
| "logps/chosen": -189.0, | |
| "logps/rejected": -164.0, | |
| "loss": 44.25, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04301364339001277, | |
| "grad_norm": 67.05587005615234, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.0625, | |
| "logps/chosen": -186.0, | |
| "logps/rejected": -156.0, | |
| "loss": 44.4067, | |
| "rewards/accuracies": 0.2100694477558136, | |
| "rewards/chosen": -0.0020751953125, | |
| "rewards/margins": -0.002960205078125, | |
| "rewards/rejected": 0.000888824462890625, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08602728678002554, | |
| "grad_norm": 56.3581428527832, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -1.046875, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -180.0, | |
| "logps/rejected": -158.0, | |
| "loss": 44.3324, | |
| "rewards/accuracies": 0.27031248807907104, | |
| "rewards/chosen": -0.000431060791015625, | |
| "rewards/margins": -0.0002727508544921875, | |
| "rewards/rejected": -0.000156402587890625, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12904093017003831, | |
| "grad_norm": 59.68115997314453, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -1.046875, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -193.0, | |
| "logps/rejected": -151.0, | |
| "loss": 44.3242, | |
| "rewards/accuracies": 0.26875001192092896, | |
| "rewards/chosen": 0.000507354736328125, | |
| "rewards/margins": -0.0001468658447265625, | |
| "rewards/rejected": 0.0006561279296875, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17205457356005108, | |
| "grad_norm": 58.759300231933594, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -186.0, | |
| "logps/rejected": -151.0, | |
| "loss": 44.2062, | |
| "rewards/accuracies": 0.2874999940395355, | |
| "rewards/chosen": 0.0030670166015625, | |
| "rewards/margins": 0.0029144287109375, | |
| "rewards/rejected": 0.0001468658447265625, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.21506821695006384, | |
| "grad_norm": 65.82920837402344, | |
| "learning_rate": 4.999361498869529e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -184.0, | |
| "logps/rejected": -158.0, | |
| "loss": 44.4336, | |
| "rewards/accuracies": 0.23906250298023224, | |
| "rewards/chosen": -0.0015869140625, | |
| "rewards/margins": -0.003143310546875, | |
| "rewards/rejected": 0.00156402587890625, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.25808186034007663, | |
| "grad_norm": 52.45038986206055, | |
| "learning_rate": 4.988019438437758e-07, | |
| "logits/chosen": -1.0390625, | |
| "logits/rejected": -1.0546875, | |
| "logps/chosen": -178.0, | |
| "logps/rejected": -159.0, | |
| "loss": 44.2992, | |
| "rewards/accuracies": 0.2671875059604645, | |
| "rewards/chosen": 0.0048828125, | |
| "rewards/margins": 0.001068115234375, | |
| "rewards/rejected": 0.003814697265625, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.30109550373008936, | |
| "grad_norm": 57.27840805053711, | |
| "learning_rate": 4.962562537324176e-07, | |
| "logits/chosen": -1.0390625, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -177.0, | |
| "logps/rejected": -158.0, | |
| "loss": 44.3105, | |
| "rewards/accuracies": 0.2750000059604645, | |
| "rewards/chosen": 0.004791259765625, | |
| "rewards/margins": 0.00021648406982421875, | |
| "rewards/rejected": 0.00457763671875, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.34410914712010215, | |
| "grad_norm": 63.51906204223633, | |
| "learning_rate": 4.923135215663896e-07, | |
| "logits/chosen": -1.03125, | |
| "logits/rejected": -1.0625, | |
| "logps/chosen": -179.0, | |
| "logps/rejected": -155.0, | |
| "loss": 44.3031, | |
| "rewards/accuracies": 0.2874999940395355, | |
| "rewards/chosen": 0.0072021484375, | |
| "rewards/margins": 0.0003147125244140625, | |
| "rewards/rejected": 0.00689697265625, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.38712279051011494, | |
| "grad_norm": 62.818267822265625, | |
| "learning_rate": 4.8699611495083e-07, | |
| "logits/chosen": -1.046875, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -181.0, | |
| "logps/rejected": -152.0, | |
| "loss": 44.2629, | |
| "rewards/accuracies": 0.3046875, | |
| "rewards/chosen": 0.0087890625, | |
| "rewards/margins": 0.002227783203125, | |
| "rewards/rejected": 0.006561279296875, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4301364339001277, | |
| "grad_norm": 62.98960494995117, | |
| "learning_rate": 4.803342001883246e-07, | |
| "logits/chosen": -1.03125, | |
| "logits/rejected": -1.0546875, | |
| "logps/chosen": -187.0, | |
| "logps/rejected": -165.0, | |
| "loss": 44.2797, | |
| "rewards/accuracies": 0.3203125, | |
| "rewards/chosen": 0.01214599609375, | |
| "rewards/margins": 0.00173187255859375, | |
| "rewards/rejected": 0.0103759765625, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4301364339001277, | |
| "eval_logits/chosen": -1.0546875, | |
| "eval_logits/rejected": -1.0703125, | |
| "eval_logps/chosen": -174.0, | |
| "eval_logps/rejected": -163.0, | |
| "eval_loss": 0.6920444369316101, | |
| "eval_rewards/accuracies": 0.32935988903045654, | |
| "eval_rewards/chosen": 0.01458740234375, | |
| "eval_rewards/margins": 0.00140380859375, | |
| "eval_rewards/rejected": 0.01318359375, | |
| "eval_runtime": 2133.8068, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 0.717, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.47315007729014047, | |
| "grad_norm": 61.54692459106445, | |
| "learning_rate": 4.72365571141757e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.0859375, | |
| "logps/chosen": -181.0, | |
| "logps/rejected": -162.0, | |
| "loss": 44.2184, | |
| "rewards/accuracies": 0.3499999940395355, | |
| "rewards/chosen": 0.015625, | |
| "rewards/margins": 0.00396728515625, | |
| "rewards/rejected": 0.01165771484375, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5161637206801533, | |
| "grad_norm": 62.887245178222656, | |
| "learning_rate": 4.6313543482507056e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.0625, | |
| "logps/chosen": -176.0, | |
| "logps/rejected": -154.0, | |
| "loss": 44.1973, | |
| "rewards/accuracies": 0.3921875059604645, | |
| "rewards/chosen": 0.02099609375, | |
| "rewards/margins": 0.0038299560546875, | |
| "rewards/rejected": 0.0172119140625, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.559177364070166, | |
| "grad_norm": 94.9865951538086, | |
| "learning_rate": 4.526961549383108e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.046875, | |
| "logps/chosen": -191.0, | |
| "logps/rejected": -158.0, | |
| "loss": 44.2, | |
| "rewards/accuracies": 0.3656249940395355, | |
| "rewards/chosen": 0.02490234375, | |
| "rewards/margins": 0.004547119140625, | |
| "rewards/rejected": 0.0203857421875, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6021910074601787, | |
| "grad_norm": 62.164451599121094, | |
| "learning_rate": 4.4110695480190597e-07, | |
| "logits/chosen": -1.078125, | |
| "logits/rejected": -1.0859375, | |
| "logps/chosen": -179.0, | |
| "logps/rejected": -153.0, | |
| "loss": 44.1516, | |
| "rewards/accuracies": 0.40937501192092896, | |
| "rewards/chosen": 0.03076171875, | |
| "rewards/margins": 0.00604248046875, | |
| "rewards/rejected": 0.024658203125, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6452046508501915, | |
| "grad_norm": 63.30311965942383, | |
| "learning_rate": 4.284335813754769e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -182.0, | |
| "logps/rejected": -158.0, | |
| "loss": 44.2184, | |
| "rewards/accuracies": 0.37968748807907104, | |
| "rewards/chosen": 0.03369140625, | |
| "rewards/margins": 0.0037689208984375, | |
| "rewards/rejected": 0.030029296875, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6882182942402043, | |
| "grad_norm": 67.73347473144531, | |
| "learning_rate": 4.1474793226723825e-07, | |
| "logits/chosen": -1.046875, | |
| "logits/rejected": -1.0546875, | |
| "logps/chosen": -172.0, | |
| "logps/rejected": -155.0, | |
| "loss": 44.1336, | |
| "rewards/accuracies": 0.4140625, | |
| "rewards/chosen": 0.04052734375, | |
| "rewards/margins": 0.006866455078125, | |
| "rewards/rejected": 0.033447265625, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7312319376302171, | |
| "grad_norm": 69.98623657226562, | |
| "learning_rate": 4.001276478500126e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -183.0, | |
| "logps/rejected": -156.0, | |
| "loss": 44.0789, | |
| "rewards/accuracies": 0.42656248807907104, | |
| "rewards/chosen": 0.044921875, | |
| "rewards/margins": 0.00726318359375, | |
| "rewards/rejected": 0.03759765625, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7742455810202299, | |
| "grad_norm": 80.5298843383789, | |
| "learning_rate": 3.846556707978337e-07, | |
| "logits/chosen": -1.0390625, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -175.0, | |
| "logps/rejected": -146.0, | |
| "loss": 43.9258, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": 0.052734375, | |
| "rewards/margins": 0.01385498046875, | |
| "rewards/rejected": 0.038818359375, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8172592244102426, | |
| "grad_norm": 72.71257781982422, | |
| "learning_rate": 3.684197755419419e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.09375, | |
| "logps/chosen": -180.0, | |
| "logps/rejected": -157.0, | |
| "loss": 43.9504, | |
| "rewards/accuracies": 0.45781248807907104, | |
| "rewards/chosen": 0.0615234375, | |
| "rewards/margins": 0.01220703125, | |
| "rewards/rejected": 0.04931640625, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8602728678002554, | |
| "grad_norm": 70.82247924804688, | |
| "learning_rate": 3.5151207031562633e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -184.0, | |
| "logps/rejected": -157.0, | |
| "loss": 44.0094, | |
| "rewards/accuracies": 0.44843751192092896, | |
| "rewards/chosen": 0.0703125, | |
| "rewards/margins": 0.01141357421875, | |
| "rewards/rejected": 0.058837890625, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8602728678002554, | |
| "eval_logits/chosen": -1.0625, | |
| "eval_logits/rejected": -1.078125, | |
| "eval_logps/chosen": -173.0, | |
| "eval_logps/rejected": -162.0, | |
| "eval_loss": 0.6888893246650696, | |
| "eval_rewards/accuracies": 0.4306009113788605, | |
| "eval_rewards/chosen": 0.0703125, | |
| "eval_rewards/margins": 0.00848388671875, | |
| "eval_rewards/rejected": 0.061767578125, | |
| "eval_runtime": 2136.7935, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 0.716, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9032865111902681, | |
| "grad_norm": 58.35607147216797, | |
| "learning_rate": 3.34028474612874e-07, | |
| "logits/chosen": -1.03125, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -184.0, | |
| "logps/rejected": -162.0, | |
| "loss": 43.9633, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": 0.07666015625, | |
| "rewards/margins": 0.01165771484375, | |
| "rewards/rejected": 0.06494140625, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9463001545802809, | |
| "grad_norm": 62.95896911621094, | |
| "learning_rate": 3.1606817502526736e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.0859375, | |
| "logps/chosen": -181.0, | |
| "logps/rejected": -158.0, | |
| "loss": 44.0883, | |
| "rewards/accuracies": 0.4234375059604645, | |
| "rewards/chosen": 0.08203125, | |
| "rewards/margins": 0.007720947265625, | |
| "rewards/rejected": 0.07421875, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9893137979702937, | |
| "grad_norm": 70.77607727050781, | |
| "learning_rate": 2.9773306254423513e-07, | |
| "logits/chosen": -1.0703125, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -186.0, | |
| "logps/rejected": -159.0, | |
| "loss": 44.0246, | |
| "rewards/accuracies": 0.4906249940395355, | |
| "rewards/chosen": 0.0888671875, | |
| "rewards/margins": 0.01190185546875, | |
| "rewards/rejected": 0.0771484375, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0323274413603065, | |
| "grad_norm": 68.3196792602539, | |
| "learning_rate": 2.791271545209101e-07, | |
| "logits/chosen": -1.0703125, | |
| "logits/rejected": -1.0859375, | |
| "logps/chosen": -190.0, | |
| "logps/rejected": -163.0, | |
| "loss": 43.8422, | |
| "rewards/accuracies": 0.4671874940395355, | |
| "rewards/chosen": 0.09716796875, | |
| "rewards/margins": 0.0157470703125, | |
| "rewards/rejected": 0.0810546875, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0753410847503193, | |
| "grad_norm": 65.2646484375, | |
| "learning_rate": 2.603560045628857e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -172.0, | |
| "logps/rejected": -152.0, | |
| "loss": 43.9477, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": 0.09521484375, | |
| "rewards/margins": 0.01373291015625, | |
| "rewards/rejected": 0.08154296875, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.118354728140332, | |
| "grad_norm": 76.79366302490234, | |
| "learning_rate": 2.4152610371560093e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.09375, | |
| "logps/chosen": -172.0, | |
| "logps/rejected": -152.0, | |
| "loss": 43.9469, | |
| "rewards/accuracies": 0.47968751192092896, | |
| "rewards/chosen": 0.10107421875, | |
| "rewards/margins": 0.01409912109375, | |
| "rewards/rejected": 0.0869140625, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.1613683715303447, | |
| "grad_norm": 62.805416107177734, | |
| "learning_rate": 2.2274427632552503e-07, | |
| "logits/chosen": -1.0859375, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -186.0, | |
| "logps/rejected": -158.0, | |
| "loss": 43.834, | |
| "rewards/accuracies": 0.47343748807907104, | |
| "rewards/chosen": 0.111328125, | |
| "rewards/margins": 0.0164794921875, | |
| "rewards/rejected": 0.09521484375, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.2043820149203575, | |
| "grad_norm": 68.53240966796875, | |
| "learning_rate": 2.0411707401248403e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -169.0, | |
| "logps/rejected": -154.0, | |
| "loss": 43.7711, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": 0.11767578125, | |
| "rewards/margins": 0.01953125, | |
| "rewards/rejected": 0.09814453125, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.2473956583103702, | |
| "grad_norm": 73.78852844238281, | |
| "learning_rate": 1.8575017118919928e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -189.0, | |
| "logps/rejected": -159.0, | |
| "loss": 43.723, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": 0.1279296875, | |
| "rewards/margins": 0.020751953125, | |
| "rewards/rejected": 0.107421875, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.290409301700383, | |
| "grad_norm": 61.855899810791016, | |
| "learning_rate": 1.6774776555733028e-07, | |
| "logits/chosen": -1.046875, | |
| "logits/rejected": -1.0859375, | |
| "logps/chosen": -175.0, | |
| "logps/rejected": -158.0, | |
| "loss": 43.8781, | |
| "rewards/accuracies": 0.49687498807907104, | |
| "rewards/chosen": 0.126953125, | |
| "rewards/margins": 0.01611328125, | |
| "rewards/rejected": 0.1103515625, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.290409301700383, | |
| "eval_logits/chosen": -1.0703125, | |
| "eval_logits/rejected": -1.0859375, | |
| "eval_logps/chosen": -171.0, | |
| "eval_logps/rejected": -161.0, | |
| "eval_loss": 0.6858682036399841, | |
| "eval_rewards/accuracies": 0.46946439146995544, | |
| "eval_rewards/chosen": 0.1328125, | |
| "eval_rewards/margins": 0.0152587890625, | |
| "eval_rewards/rejected": 0.11767578125, | |
| "eval_runtime": 2134.9779, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 0.717, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.3334229450903958, | |
| "grad_norm": 67.3431625366211, | |
| "learning_rate": 1.5021198698108036e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -190.0, | |
| "logps/rejected": -166.0, | |
| "loss": 43.9168, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": 0.1396484375, | |
| "rewards/margins": 0.01483154296875, | |
| "rewards/rejected": 0.12451171875, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.3764365884804086, | |
| "grad_norm": 72.8765640258789, | |
| "learning_rate": 1.3324231809189983e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -190.0, | |
| "logps/rejected": -165.0, | |
| "loss": 43.6699, | |
| "rewards/accuracies": 0.520312488079071, | |
| "rewards/chosen": 0.146484375, | |
| "rewards/margins": 0.023193359375, | |
| "rewards/rejected": 0.123046875, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.4194502318704214, | |
| "grad_norm": 66.81339263916016, | |
| "learning_rate": 1.1693502991126608e-07, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.109375, | |
| "logps/chosen": -181.0, | |
| "logps/rejected": -152.0, | |
| "loss": 43.5453, | |
| "rewards/accuracies": 0.5015624761581421, | |
| "rewards/chosen": 0.1474609375, | |
| "rewards/margins": 0.027099609375, | |
| "rewards/rejected": 0.12060546875, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.4624638752604342, | |
| "grad_norm": 61.55015563964844, | |
| "learning_rate": 1.0138263569332267e-07, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.0703125, | |
| "logps/chosen": -186.0, | |
| "logps/rejected": -152.0, | |
| "loss": 43.5445, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.15234375, | |
| "rewards/margins": 0.0263671875, | |
| "rewards/rejected": 0.1259765625, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.505477518650447, | |
| "grad_norm": 66.9729232788086, | |
| "learning_rate": 8.667336608579487e-08, | |
| "logits/chosen": -1.078125, | |
| "logits/rejected": -1.1015625, | |
| "logps/chosen": -175.0, | |
| "logps/rejected": -151.0, | |
| "loss": 43.4664, | |
| "rewards/accuracies": 0.5328124761581421, | |
| "rewards/chosen": 0.1552734375, | |
| "rewards/margins": 0.0291748046875, | |
| "rewards/rejected": 0.1259765625, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.5484911620404596, | |
| "grad_norm": 60.46580505371094, | |
| "learning_rate": 7.28906685866599e-08, | |
| "logits/chosen": -1.078125, | |
| "logits/rejected": -1.1015625, | |
| "logps/chosen": -180.0, | |
| "logps/rejected": -160.0, | |
| "loss": 43.6453, | |
| "rewards/accuracies": 0.5218750238418579, | |
| "rewards/chosen": 0.1552734375, | |
| "rewards/margins": 0.022705078125, | |
| "rewards/rejected": 0.1318359375, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.5915048054304726, | |
| "grad_norm": 63.649444580078125, | |
| "learning_rate": 6.01127341362138e-08, | |
| "logits/chosen": -1.078125, | |
| "logits/rejected": -1.1015625, | |
| "logps/chosen": -172.0, | |
| "logps/rejected": -151.0, | |
| "loss": 43.7188, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": 0.154296875, | |
| "rewards/margins": 0.02197265625, | |
| "rewards/rejected": 0.1328125, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.6345184488204851, | |
| "grad_norm": 64.85546875, | |
| "learning_rate": 4.841205353023714e-08, | |
| "logits/chosen": -1.0703125, | |
| "logits/rejected": -1.09375, | |
| "logps/chosen": -173.0, | |
| "logps/rejected": -155.0, | |
| "loss": 43.8621, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": 0.154296875, | |
| "rewards/margins": 0.018310546875, | |
| "rewards/rejected": 0.1357421875, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.6775320922104981, | |
| "grad_norm": 65.72016143798828, | |
| "learning_rate": 3.785500617078424e-08, | |
| "logits/chosen": -1.0625, | |
| "logits/rejected": -1.078125, | |
| "logps/chosen": -177.0, | |
| "logps/rejected": -150.0, | |
| "loss": 43.6246, | |
| "rewards/accuracies": 0.5140625238418579, | |
| "rewards/chosen": 0.1552734375, | |
| "rewards/margins": 0.0245361328125, | |
| "rewards/rejected": 0.130859375, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.7205457356005107, | |
| "grad_norm": 61.353736877441406, | |
| "learning_rate": 2.850148348765921e-08, | |
| "logits/chosen": -1.0703125, | |
| "logits/rejected": -1.1015625, | |
| "logps/chosen": -171.0, | |
| "logps/rejected": -150.0, | |
| "loss": 43.7996, | |
| "rewards/accuracies": 0.49531251192092896, | |
| "rewards/chosen": 0.15625, | |
| "rewards/margins": 0.018310546875, | |
| "rewards/rejected": 0.138671875, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.7205457356005107, | |
| "eval_logits/chosen": -1.0703125, | |
| "eval_logits/rejected": -1.0859375, | |
| "eval_logps/chosen": -171.0, | |
| "eval_logps/rejected": -161.0, | |
| "eval_loss": 0.6854622960090637, | |
| "eval_rewards/accuracies": 0.48187458515167236, | |
| "eval_rewards/chosen": 0.1552734375, | |
| "eval_rewards/margins": 0.0166015625, | |
| "eval_rewards/rejected": 0.138671875, | |
| "eval_runtime": 2129.0581, | |
| "eval_samples_per_second": 2.875, | |
| "eval_steps_per_second": 0.719, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.7635593789905235, | |
| "grad_norm": 67.52031707763672, | |
| "learning_rate": 2.0404549166959718e-08, | |
| "logits/chosen": -1.078125, | |
| "logits/rejected": -1.1015625, | |
| "logps/chosen": -183.0, | |
| "logps/rejected": -149.0, | |
| "loss": 43.5926, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": 0.1611328125, | |
| "rewards/margins": 0.0257568359375, | |
| "rewards/rejected": 0.134765625, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.8065730223805363, | |
| "grad_norm": 63.740577697753906, | |
| "learning_rate": 1.3610138114250519e-08, | |
| "logits/chosen": -1.0703125, | |
| "logits/rejected": -1.1015625, | |
| "logps/chosen": -174.0, | |
| "logps/rejected": -155.0, | |
| "loss": 43.6965, | |
| "rewards/accuracies": 0.4859375059604645, | |
| "rewards/chosen": 0.158203125, | |
| "rewards/margins": 0.0225830078125, | |
| "rewards/rejected": 0.1357421875, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.849586665770549, | |
| "grad_norm": 65.15459442138672, | |
| "learning_rate": 8.156795860187027e-09, | |
| "logits/chosen": -1.078125, | |
| "logits/rejected": -1.1171875, | |
| "logps/chosen": -184.0, | |
| "logps/rejected": -157.0, | |
| "loss": 43.6516, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": 0.1591796875, | |
| "rewards/margins": 0.0234375, | |
| "rewards/rejected": 0.1357421875, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.8926003091605619, | |
| "grad_norm": 67.16203308105469, | |
| "learning_rate": 4.075459886973082e-09, | |
| "logits/chosen": -1.0546875, | |
| "logits/rejected": -1.09375, | |
| "logps/chosen": -179.0, | |
| "logps/rejected": -145.0, | |
| "loss": 43.5895, | |
| "rewards/accuracies": 0.5171874761581421, | |
| "rewards/chosen": 0.1611328125, | |
| "rewards/margins": 0.0262451171875, | |
| "rewards/rejected": 0.134765625, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.9356139525505747, | |
| "grad_norm": 64.06181335449219, | |
| "learning_rate": 1.3892841162143899e-09, | |
| "logits/chosen": -1.078125, | |
| "logits/rejected": -1.1015625, | |
| "logps/chosen": -183.0, | |
| "logps/rejected": -149.0, | |
| "loss": 43.5109, | |
| "rewards/accuracies": 0.526562511920929, | |
| "rewards/chosen": 0.162109375, | |
| "rewards/margins": 0.0274658203125, | |
| "rewards/rejected": 0.134765625, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.9786275959405875, | |
| "grad_norm": 71.79183959960938, | |
| "learning_rate": 1.1350755386951849e-10, | |
| "logits/chosen": -1.0859375, | |
| "logits/rejected": -1.1171875, | |
| "logps/chosen": -182.0, | |
| "logps/rejected": -155.0, | |
| "loss": 43.7199, | |
| "rewards/accuracies": 0.5328124761581421, | |
| "rewards/chosen": 0.1591796875, | |
| "rewards/margins": 0.0223388671875, | |
| "rewards/rejected": 0.13671875, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.9958330532965927, | |
| "step": 464, | |
| "total_flos": 0.0, | |
| "train_loss": 43.948444234913794, | |
| "train_runtime": 23482.907, | |
| "train_samples_per_second": 1.267, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 464, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |