| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 939, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003194888178913738, | |
| "grad_norm": 66.3561626483212, | |
| "learning_rate": 5.3191489361702125e-09, | |
| "logits/chosen": -1.6796875, | |
| "logits/rejected": -2.546875, | |
| "logps/chosen": -302.0, | |
| "logps/rejected": -145.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03194888178913738, | |
| "grad_norm": 67.57820401478659, | |
| "learning_rate": 5.3191489361702123e-08, | |
| "logits/chosen": -1.6640625, | |
| "logits/rejected": -2.5625, | |
| "logps/chosen": -316.0, | |
| "logps/rejected": -153.0, | |
| "loss": 0.6886, | |
| "rewards/accuracies": 0.1388888955116272, | |
| "rewards/chosen": 0.0111083984375, | |
| "rewards/margins": -0.0006866455078125, | |
| "rewards/rejected": 0.01177978515625, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06389776357827476, | |
| "grad_norm": 59.43443097233231, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.6640625, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -312.0, | |
| "logps/rejected": -157.0, | |
| "loss": 0.6598, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.0225830078125, | |
| "rewards/margins": 0.06787109375, | |
| "rewards/rejected": -0.045166015625, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 42.18507123666234, | |
| "learning_rate": 1.5957446808510638e-07, | |
| "logits/chosen": -1.6640625, | |
| "logits/rejected": -2.40625, | |
| "logps/chosen": -310.0, | |
| "logps/rejected": -158.0, | |
| "loss": 0.5472, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.06005859375, | |
| "rewards/margins": 0.330078125, | |
| "rewards/rejected": -0.26953125, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12779552715654952, | |
| "grad_norm": 18.160729456486774, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -1.65625, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -312.0, | |
| "logps/rejected": -166.0, | |
| "loss": 0.3235, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0673828125, | |
| "rewards/margins": 1.0625, | |
| "rewards/rejected": -0.99609375, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1597444089456869, | |
| "grad_norm": 6.161021791324783, | |
| "learning_rate": 2.659574468085106e-07, | |
| "logits/chosen": -1.6640625, | |
| "logits/rejected": -2.390625, | |
| "logps/chosen": -316.0, | |
| "logps/rejected": -177.0, | |
| "loss": 0.0931, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.25, | |
| "rewards/margins": 2.5, | |
| "rewards/rejected": -2.25, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 1.4155154630049935, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -314.0, | |
| "logps/rejected": -189.0, | |
| "loss": 0.0225, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.546875, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -3.671875, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22364217252396165, | |
| "grad_norm": 0.8420950404007327, | |
| "learning_rate": 3.7234042553191484e-07, | |
| "logits/chosen": -1.6796875, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -205.0, | |
| "loss": 0.0034, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.078125, | |
| "rewards/margins": 6.25, | |
| "rewards/rejected": -5.1875, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25559105431309903, | |
| "grad_norm": 0.24119281873509665, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -1.5, | |
| "logits/rejected": -2.390625, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -215.0, | |
| "loss": 0.0012, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.4453125, | |
| "rewards/margins": 7.34375, | |
| "rewards/rejected": -5.875, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 0.055464793372892944, | |
| "learning_rate": 4.787234042553192e-07, | |
| "logits/chosen": -1.6328125, | |
| "logits/rejected": -2.375, | |
| "logps/chosen": -306.0, | |
| "logps/rejected": -217.0, | |
| "loss": 0.0007, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.453125, | |
| "rewards/margins": 7.75, | |
| "rewards/rejected": -6.3125, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3194888178913738, | |
| "grad_norm": 0.05859296499933451, | |
| "learning_rate": 4.964497041420119e-07, | |
| "logits/chosen": -1.6484375, | |
| "logits/rejected": -2.375, | |
| "logps/chosen": -300.0, | |
| "logps/rejected": -219.0, | |
| "loss": 0.0007, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.5546875, | |
| "rewards/margins": 8.25, | |
| "rewards/rejected": -6.65625, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3514376996805112, | |
| "grad_norm": 0.08102356000051511, | |
| "learning_rate": 4.905325443786982e-07, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -300.0, | |
| "logps/rejected": -225.0, | |
| "loss": 0.0005, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.515625, | |
| "rewards/margins": 8.5625, | |
| "rewards/rejected": -7.0625, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 0.042779195932783155, | |
| "learning_rate": 4.846153846153846e-07, | |
| "logits/chosen": -1.6484375, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -233.0, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.8515625, | |
| "rewards/margins": 9.1875, | |
| "rewards/rejected": -7.34375, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.41533546325878595, | |
| "grad_norm": 0.13985054295121171, | |
| "learning_rate": 4.78698224852071e-07, | |
| "logits/chosen": -1.6328125, | |
| "logits/rejected": -2.328125, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -239.0, | |
| "loss": 0.0005, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.75, | |
| "rewards/margins": 9.1875, | |
| "rewards/rejected": -7.46875, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4472843450479233, | |
| "grad_norm": 0.3290125444518827, | |
| "learning_rate": 4.727810650887574e-07, | |
| "logits/chosen": -1.640625, | |
| "logits/rejected": -2.546875, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -233.0, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.7265625, | |
| "rewards/margins": 9.5625, | |
| "rewards/rejected": -7.84375, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 0.008495508515545779, | |
| "learning_rate": 4.668639053254438e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.578125, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -231.0, | |
| "loss": 0.0017, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.84375, | |
| "rewards/margins": 9.75, | |
| "rewards/rejected": -7.9375, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5111821086261981, | |
| "grad_norm": 0.2101365543875086, | |
| "learning_rate": 4.6094674556213014e-07, | |
| "logits/chosen": -1.6015625, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -233.0, | |
| "loss": 0.0002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.8046875, | |
| "rewards/margins": 9.6875, | |
| "rewards/rejected": -7.875, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5431309904153354, | |
| "grad_norm": 0.014836072743792679, | |
| "learning_rate": 4.5502958579881655e-07, | |
| "logits/chosen": -1.6484375, | |
| "logits/rejected": -2.5625, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -233.0, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.8046875, | |
| "rewards/margins": 9.9375, | |
| "rewards/rejected": -8.125, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 0.010891947734143552, | |
| "learning_rate": 4.491124260355029e-07, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -242.0, | |
| "loss": 0.0002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.78125, | |
| "rewards/margins": 10.125, | |
| "rewards/rejected": -8.3125, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6070287539936102, | |
| "grad_norm": 0.05412735956611531, | |
| "learning_rate": 4.4319526627218936e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -302.0, | |
| "logps/rejected": -240.0, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.8359375, | |
| "rewards/margins": 10.3125, | |
| "rewards/rejected": -8.5, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6389776357827476, | |
| "grad_norm": 0.027506697448395234, | |
| "learning_rate": 4.372781065088757e-07, | |
| "logits/chosen": -1.6796875, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -304.0, | |
| "logps/rejected": -244.0, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.8203125, | |
| "rewards/margins": 10.3125, | |
| "rewards/rejected": -8.5, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 0.0060678459132032575, | |
| "learning_rate": 4.313609467455621e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -300.0, | |
| "logps/rejected": -235.0, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.0, | |
| "rewards/margins": 10.5, | |
| "rewards/rejected": -8.5, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7028753993610224, | |
| "grad_norm": 0.006506686204652437, | |
| "learning_rate": 4.2544378698224847e-07, | |
| "logits/chosen": -1.6796875, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -300.0, | |
| "logps/rejected": -247.0, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.9765625, | |
| "rewards/margins": 10.6875, | |
| "rewards/rejected": -8.6875, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7348242811501597, | |
| "grad_norm": 0.007031528224055077, | |
| "learning_rate": 4.195266272189349e-07, | |
| "logits/chosen": -1.671875, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -245.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.9609375, | |
| "rewards/margins": 10.8125, | |
| "rewards/rejected": -8.8125, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 0.003985253197057443, | |
| "learning_rate": 4.1360946745562133e-07, | |
| "logits/chosen": -1.640625, | |
| "logits/rejected": -2.53125, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -238.0, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.15625, | |
| "rewards/margins": 10.75, | |
| "rewards/rejected": -8.625, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7987220447284346, | |
| "grad_norm": 0.005795902810172554, | |
| "learning_rate": 4.076923076923077e-07, | |
| "logits/chosen": -1.6484375, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -247.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.0625, | |
| "rewards/margins": 10.875, | |
| "rewards/rejected": -8.8125, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8306709265175719, | |
| "grad_norm": 0.004800183675266446, | |
| "learning_rate": 4.017751479289941e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -248.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.109375, | |
| "rewards/margins": 11.0625, | |
| "rewards/rejected": -8.9375, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 0.005883343730949623, | |
| "learning_rate": 3.9585798816568044e-07, | |
| "logits/chosen": -1.6328125, | |
| "logits/rejected": -2.546875, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -242.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.0625, | |
| "rewards/margins": 10.9375, | |
| "rewards/rejected": -8.875, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8945686900958466, | |
| "grad_norm": 0.003942793783714264, | |
| "learning_rate": 3.8994082840236685e-07, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -2.53125, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -243.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.046875, | |
| "rewards/margins": 10.9375, | |
| "rewards/rejected": -8.875, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9265175718849841, | |
| "grad_norm": 0.004433194446356253, | |
| "learning_rate": 3.840236686390532e-07, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -2.390625, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -241.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.09375, | |
| "rewards/margins": 11.0, | |
| "rewards/rejected": -8.875, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.0030444166246834195, | |
| "learning_rate": 3.7810650887573966e-07, | |
| "logits/chosen": -1.5703125, | |
| "logits/rejected": -2.53125, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -242.0, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.890625, | |
| "rewards/margins": 11.125, | |
| "rewards/rejected": -9.25, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9904153354632588, | |
| "grad_norm": 0.003966733414361215, | |
| "learning_rate": 3.72189349112426e-07, | |
| "logits/chosen": -1.65625, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -246.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.15625, | |
| "rewards/margins": 11.4375, | |
| "rewards/rejected": -9.25, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_logits/chosen": -1.59375, | |
| "eval_logits/rejected": -2.53125, | |
| "eval_logps/chosen": -292.0, | |
| "eval_logps/rejected": -244.0, | |
| "eval_loss": 2.0532608687062748e-05, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 2.09375, | |
| "eval_rewards/margins": 11.5625, | |
| "eval_rewards/rejected": -9.4375, | |
| "eval_runtime": 10.5272, | |
| "eval_samples_per_second": 18.998, | |
| "eval_steps_per_second": 0.665, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.0223642172523961, | |
| "grad_norm": 0.002558243370335936, | |
| "learning_rate": 3.662721893491124e-07, | |
| "logits/chosen": -1.546875, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -251.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.125, | |
| "rewards/margins": 11.5, | |
| "rewards/rejected": -9.375, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 0.0021881003082927803, | |
| "learning_rate": 3.603550295857988e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -251.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.109375, | |
| "rewards/margins": 11.4375, | |
| "rewards/rejected": -9.375, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0862619808306708, | |
| "grad_norm": 0.0020626605312609597, | |
| "learning_rate": 3.544378698224852e-07, | |
| "logits/chosen": -1.6015625, | |
| "logits/rejected": -2.59375, | |
| "logps/chosen": -286.0, | |
| "logps/rejected": -251.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.125, | |
| "rewards/margins": 11.5625, | |
| "rewards/rejected": -9.4375, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1182108626198084, | |
| "grad_norm": 0.0039043682638062066, | |
| "learning_rate": 3.485207100591716e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -252.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.109375, | |
| "rewards/margins": 11.375, | |
| "rewards/rejected": -9.25, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.0034990794114765538, | |
| "learning_rate": 3.42603550295858e-07, | |
| "logits/chosen": -1.6640625, | |
| "logits/rejected": -2.328125, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -252.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.109375, | |
| "rewards/margins": 11.4375, | |
| "rewards/rejected": -9.3125, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.182108626198083, | |
| "grad_norm": 0.002380656478840785, | |
| "learning_rate": 3.366863905325444e-07, | |
| "logits/chosen": -1.6015625, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -254.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.125, | |
| "rewards/margins": 11.5625, | |
| "rewards/rejected": -9.375, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2140575079872205, | |
| "grad_norm": 0.0013940482783028198, | |
| "learning_rate": 3.3076923076923075e-07, | |
| "logits/chosen": -1.65625, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -254.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.15625, | |
| "rewards/margins": 11.75, | |
| "rewards/rejected": -9.5625, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 0.002506190095368227, | |
| "learning_rate": 3.2485207100591716e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -253.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.15625, | |
| "rewards/margins": 11.75, | |
| "rewards/rejected": -9.5625, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2779552715654952, | |
| "grad_norm": 0.018231014973408444, | |
| "learning_rate": 3.189349112426035e-07, | |
| "logits/chosen": -1.65625, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -304.0, | |
| "logps/rejected": -245.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.96875, | |
| "rewards/margins": 11.6875, | |
| "rewards/rejected": -9.6875, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3099041533546325, | |
| "grad_norm": 0.0023577978609173114, | |
| "learning_rate": 3.130177514792899e-07, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -255.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.15625, | |
| "rewards/margins": 11.6875, | |
| "rewards/rejected": -9.5, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.0016312758047640993, | |
| "learning_rate": 3.071005917159763e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -250.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.1875, | |
| "rewards/margins": 12.0625, | |
| "rewards/rejected": -9.875, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3738019169329074, | |
| "grad_norm": 0.002962282707083474, | |
| "learning_rate": 3.011834319526627e-07, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -245.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.28125, | |
| "rewards/margins": 11.75, | |
| "rewards/rejected": -9.5, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4057507987220448, | |
| "grad_norm": 0.0020582153728422687, | |
| "learning_rate": 2.952662721893491e-07, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -252.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.203125, | |
| "rewards/margins": 12.0625, | |
| "rewards/rejected": -9.875, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 0.0015864014541557795, | |
| "learning_rate": 2.893491124260355e-07, | |
| "logits/chosen": -1.5390625, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -252.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.265625, | |
| "rewards/margins": 12.0, | |
| "rewards/rejected": -9.75, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4696485623003195, | |
| "grad_norm": 0.0014107244301989241, | |
| "learning_rate": 2.834319526627219e-07, | |
| "logits/chosen": -1.671875, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -300.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.171875, | |
| "rewards/margins": 12.125, | |
| "rewards/rejected": -9.9375, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5015974440894568, | |
| "grad_norm": 0.0015455009357666874, | |
| "learning_rate": 2.7751479289940824e-07, | |
| "logits/chosen": -1.5703125, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.1875, | |
| "rewards/margins": 12.125, | |
| "rewards/rejected": -9.9375, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.021325167782399665, | |
| "learning_rate": 2.715976331360947e-07, | |
| "logits/chosen": -1.5546875, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -256.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.203125, | |
| "rewards/margins": 12.0, | |
| "rewards/rejected": -9.8125, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5654952076677318, | |
| "grad_norm": 0.00191440770150257, | |
| "learning_rate": 2.6568047337278106e-07, | |
| "logits/chosen": -1.7265625, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -306.0, | |
| "logps/rejected": -254.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.140625, | |
| "rewards/margins": 12.125, | |
| "rewards/rejected": -10.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.5974440894568689, | |
| "grad_norm": 0.0015995149381860404, | |
| "learning_rate": 2.5976331360946746e-07, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -2.359375, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -254.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.25, | |
| "rewards/margins": 12.1875, | |
| "rewards/rejected": -9.9375, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 0.003998439375448258, | |
| "learning_rate": 2.538461538461538e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -256.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.3125, | |
| "rewards/margins": 12.1875, | |
| "rewards/rejected": -9.875, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6613418530351438, | |
| "grad_norm": 0.00195622435888153, | |
| "learning_rate": 2.479289940828402e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -256.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.234375, | |
| "rewards/margins": 12.3125, | |
| "rewards/rejected": -10.0625, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.6932907348242812, | |
| "grad_norm": 0.0009779108939817446, | |
| "learning_rate": 2.420118343195266e-07, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.34375, | |
| "rewards/margins": 12.5625, | |
| "rewards/rejected": -10.25, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.00159015941098779, | |
| "learning_rate": 2.36094674556213e-07, | |
| "logits/chosen": -1.6875, | |
| "logits/rejected": -2.546875, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -254.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.203125, | |
| "rewards/margins": 12.4375, | |
| "rewards/rejected": -10.25, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.7571884984025559, | |
| "grad_norm": 0.0016071770387580436, | |
| "learning_rate": 2.301775147928994e-07, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -255.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.296875, | |
| "rewards/margins": 12.5, | |
| "rewards/rejected": -10.125, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7891373801916934, | |
| "grad_norm": 0.006676398501698908, | |
| "learning_rate": 2.242603550295858e-07, | |
| "logits/chosen": -1.65625, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -300.0, | |
| "logps/rejected": -258.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.1875, | |
| "rewards/margins": 12.3125, | |
| "rewards/rejected": -10.125, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 0.001241193021607329, | |
| "learning_rate": 2.1834319526627217e-07, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -256.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.3125, | |
| "rewards/margins": 12.625, | |
| "rewards/rejected": -10.3125, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8530351437699681, | |
| "grad_norm": 0.000630211357395811, | |
| "learning_rate": 2.1242603550295858e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.21875, | |
| "rewards/margins": 12.625, | |
| "rewards/rejected": -10.4375, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.8849840255591053, | |
| "grad_norm": 0.0007502946953630078, | |
| "learning_rate": 2.0650887573964496e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -258.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.3125, | |
| "rewards/margins": 12.625, | |
| "rewards/rejected": -10.3125, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.0008324781976188686, | |
| "learning_rate": 2.0059171597633133e-07, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -262.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.203125, | |
| "rewards/margins": 12.75, | |
| "rewards/rejected": -10.5625, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9488817891373802, | |
| "grad_norm": 0.0007778777435710634, | |
| "learning_rate": 1.9467455621301774e-07, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.34375, | |
| "rewards/margins": 12.8125, | |
| "rewards/rejected": -10.5, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.9808306709265175, | |
| "grad_norm": 0.0005882280022083605, | |
| "learning_rate": 1.8875739644970412e-07, | |
| "logits/chosen": -1.6328125, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.21875, | |
| "rewards/margins": 12.6875, | |
| "rewards/rejected": -10.5, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_logits/chosen": -1.5859375, | |
| "eval_logits/rejected": -2.53125, | |
| "eval_logps/chosen": -290.0, | |
| "eval_logps/rejected": -255.0, | |
| "eval_loss": 5.943179075984517e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 2.328125, | |
| "eval_rewards/margins": 12.9375, | |
| "eval_rewards/rejected": -10.625, | |
| "eval_runtime": 10.5621, | |
| "eval_samples_per_second": 18.936, | |
| "eval_steps_per_second": 0.663, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.012779552715655, | |
| "grad_norm": 0.0005002851689091564, | |
| "learning_rate": 1.8284023668639053e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.53125, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -262.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.203125, | |
| "rewards/margins": 12.875, | |
| "rewards/rejected": -10.625, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0447284345047922, | |
| "grad_norm": 0.0004935955721736502, | |
| "learning_rate": 1.7692307692307693e-07, | |
| "logits/chosen": -1.6640625, | |
| "logits/rejected": -2.625, | |
| "logps/chosen": -300.0, | |
| "logps/rejected": -268.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.4375, | |
| "rewards/margins": 13.0, | |
| "rewards/rejected": -10.625, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.07667731629393, | |
| "grad_norm": 0.0007716999750143461, | |
| "learning_rate": 1.710059171597633e-07, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -258.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.296875, | |
| "rewards/margins": 12.875, | |
| "rewards/rejected": -10.5625, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.0005068386241484783, | |
| "learning_rate": 1.650887573964497e-07, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -2.375, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -262.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.296875, | |
| "rewards/margins": 12.875, | |
| "rewards/rejected": -10.625, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.1405750798722045, | |
| "grad_norm": 0.000981502457343838, | |
| "learning_rate": 1.591715976331361e-07, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.296875, | |
| "rewards/margins": 12.875, | |
| "rewards/rejected": -10.5625, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.1725239616613417, | |
| "grad_norm": 0.0006440614902733118, | |
| "learning_rate": 1.5325443786982248e-07, | |
| "logits/chosen": -1.5703125, | |
| "logits/rejected": -2.546875, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.40625, | |
| "rewards/margins": 12.9375, | |
| "rewards/rejected": -10.5, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.2044728434504792, | |
| "grad_norm": 0.0005968138344014539, | |
| "learning_rate": 1.4733727810650885e-07, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.5625, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -258.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.328125, | |
| "rewards/margins": 13.0625, | |
| "rewards/rejected": -10.75, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.236421725239617, | |
| "grad_norm": 0.0008891475107269685, | |
| "learning_rate": 1.4142011834319526e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.28125, | |
| "rewards/margins": 13.125, | |
| "rewards/rejected": -10.8125, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.268370607028754, | |
| "grad_norm": 0.0004374380373679923, | |
| "learning_rate": 1.3550295857988164e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.328125, | |
| "rewards/margins": 12.75, | |
| "rewards/rejected": -10.4375, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.01662398838742337, | |
| "learning_rate": 1.2958579881656802e-07, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.484375, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.390625, | |
| "rewards/margins": 13.25, | |
| "rewards/rejected": -10.8125, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3322683706070286, | |
| "grad_norm": 0.0005505191887041714, | |
| "learning_rate": 1.2366863905325443e-07, | |
| "logits/chosen": -1.5703125, | |
| "logits/rejected": -2.546875, | |
| "logps/chosen": -286.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.359375, | |
| "rewards/margins": 13.0625, | |
| "rewards/rejected": -10.6875, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.364217252396166, | |
| "grad_norm": 0.0010480950420334609, | |
| "learning_rate": 1.1775147928994082e-07, | |
| "logits/chosen": -1.6484375, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -266.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.25, | |
| "rewards/margins": 13.0, | |
| "rewards/rejected": -10.75, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.3961661341853033, | |
| "grad_norm": 0.00043727637963375055, | |
| "learning_rate": 1.1183431952662721e-07, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.375, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -270.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.25, | |
| "rewards/margins": 13.0, | |
| "rewards/rejected": -10.6875, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.428115015974441, | |
| "grad_norm": 0.000563995140849502, | |
| "learning_rate": 1.059171597633136e-07, | |
| "logits/chosen": -1.5390625, | |
| "logits/rejected": -2.546875, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -260.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.3125, | |
| "rewards/margins": 13.1875, | |
| "rewards/rejected": -10.875, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.460063897763578, | |
| "grad_norm": 0.0004968795273638076, | |
| "learning_rate": 1e-07, | |
| "logits/chosen": -1.5546875, | |
| "logits/rejected": -2.375, | |
| "logps/chosen": -286.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.34375, | |
| "rewards/margins": 13.3125, | |
| "rewards/rejected": -11.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.0005720975269378952, | |
| "learning_rate": 9.408284023668639e-08, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -266.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.359375, | |
| "rewards/margins": 13.1875, | |
| "rewards/rejected": -10.8125, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.523961661341853, | |
| "grad_norm": 0.0005327187414650777, | |
| "learning_rate": 8.816568047337278e-08, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.390625, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -258.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.421875, | |
| "rewards/margins": 13.1875, | |
| "rewards/rejected": -10.75, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.5559105431309903, | |
| "grad_norm": 0.005188839616423619, | |
| "learning_rate": 8.224852071005916e-08, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.296875, | |
| "rewards/margins": 13.25, | |
| "rewards/rejected": -11.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.587859424920128, | |
| "grad_norm": 0.0016030087833627544, | |
| "learning_rate": 7.633136094674555e-08, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -290.0, | |
| "logps/rejected": -258.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.3125, | |
| "rewards/margins": 13.0, | |
| "rewards/rejected": -10.75, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.619808306709265, | |
| "grad_norm": 0.00046525532881776, | |
| "learning_rate": 7.041420118343195e-08, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -2.453125, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -266.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.234375, | |
| "rewards/margins": 13.0625, | |
| "rewards/rejected": -10.8125, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.6517571884984026, | |
| "grad_norm": 0.00047364005580776986, | |
| "learning_rate": 6.449704142011835e-08, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.53125, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.359375, | |
| "rewards/margins": 13.375, | |
| "rewards/rejected": -11.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.0006053041410896476, | |
| "learning_rate": 5.857988165680473e-08, | |
| "logits/chosen": -1.640625, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -266.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.28125, | |
| "rewards/margins": 13.3125, | |
| "rewards/rejected": -11.0625, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.7156549520766773, | |
| "grad_norm": 0.0008894808262520896, | |
| "learning_rate": 5.266272189349112e-08, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -266.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.25, | |
| "rewards/margins": 13.375, | |
| "rewards/rejected": -11.125, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.747603833865815, | |
| "grad_norm": 0.0003225485997448219, | |
| "learning_rate": 4.674556213017751e-08, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -286.0, | |
| "logps/rejected": -272.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.296875, | |
| "rewards/margins": 13.4375, | |
| "rewards/rejected": -11.125, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.779552715654952, | |
| "grad_norm": 0.0005331001939899559, | |
| "learning_rate": 4.082840236686391e-08, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -2.4375, | |
| "logps/chosen": -288.0, | |
| "logps/rejected": -270.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.421875, | |
| "rewards/margins": 13.4375, | |
| "rewards/rejected": -11.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.8115015974440896, | |
| "grad_norm": 0.00045862156551760734, | |
| "learning_rate": 3.4911242603550294e-08, | |
| "logits/chosen": -1.609375, | |
| "logits/rejected": -2.53125, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -262.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.328125, | |
| "rewards/margins": 13.3125, | |
| "rewards/rejected": -11.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.8434504792332267, | |
| "grad_norm": 0.0003648218080749694, | |
| "learning_rate": 2.8994082840236687e-08, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -2.53125, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -266.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.34375, | |
| "rewards/margins": 13.375, | |
| "rewards/rejected": -11.0625, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.0004039575345222897, | |
| "learning_rate": 2.3076923076923076e-08, | |
| "logits/chosen": -1.640625, | |
| "logits/rejected": -2.46875, | |
| "logps/chosen": -296.0, | |
| "logps/rejected": -266.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.375, | |
| "rewards/margins": 13.5, | |
| "rewards/rejected": -11.125, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.9073482428115014, | |
| "grad_norm": 0.0003976188210629937, | |
| "learning_rate": 1.7159763313609465e-08, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -2.375, | |
| "logps/chosen": -298.0, | |
| "logps/rejected": -268.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.359375, | |
| "rewards/margins": 13.3125, | |
| "rewards/rejected": -10.9375, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.939297124600639, | |
| "grad_norm": 0.0002899245955406109, | |
| "learning_rate": 1.1242603550295858e-08, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -294.0, | |
| "logps/rejected": -270.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.46875, | |
| "rewards/margins": 13.75, | |
| "rewards/rejected": -11.25, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.9712460063897765, | |
| "grad_norm": 0.0011828130970183842, | |
| "learning_rate": 5.325443786982248e-09, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -2.59375, | |
| "logps/chosen": -292.0, | |
| "logps/rejected": -264.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.265625, | |
| "rewards/margins": 13.5625, | |
| "rewards/rejected": -11.3125, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_logits/chosen": -1.578125, | |
| "eval_logits/rejected": -2.53125, | |
| "eval_logps/chosen": -290.0, | |
| "eval_logps/rejected": -262.0, | |
| "eval_loss": 3.544539140420966e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 2.390625, | |
| "eval_rewards/margins": 13.625, | |
| "eval_rewards/rejected": -11.25, | |
| "eval_runtime": 13.2531, | |
| "eval_samples_per_second": 15.091, | |
| "eval_steps_per_second": 0.528, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 939, | |
| "total_flos": 0.0, | |
| "train_loss": 0.024984648726936093, | |
| "train_runtime": 4521.6379, | |
| "train_samples_per_second": 6.635, | |
| "train_steps_per_second": 0.208 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 939, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |