Training in progress, step 20

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b58432df35bd04b3e5634551b7d1d2dd114fc40010499996400d1d38b4998666
+oid sha256:9c888db2569987e13d1c7cbbad0312c3a9366265d12698fc8268718d372c0d68
 size 6627156248

logging.jsonl

@@ -10,3 +10,14 @@
 {"loss": 0.54727173, "grad_norm": 0.13393623, "learning_rate": 3.88e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004846, "rewards/chosen": -0.03527832, "rewards/rejected": -0.49353027, "rewards/accuracies": 0.25, "rewards/margins": 0.45831299, "logps/rejected": -4.92724609, "logps/chosen": -0.35327148, "logits/rejected": -0.56445312, "logits/chosen": -0.82910156, "nll_loss": 0.35327148, "log_odds_ratio": -0.7265625, "log_odds_chosen": 4.79394531, "epoch": 0.0234192, "global_step/max_steps": "10/1281", "percentage": "0.78%", "elapsed_time": "34m 8s", "remaining_time": "3d 0h 18m 50s"}
 {"eval_loss": 0.66295749, "eval_runtime": 642.7006, "eval_samples_per_second": 0.657, "eval_steps_per_second": 0.082, "eval_rewards/chosen": -0.05507674, "eval_rewards/rejected": -0.09295309, "eval_rewards/accuracies": 0.1509434, "eval_rewards/margins": 0.03781186, "eval_logps/rejected": -0.92890441, "eval_logps/chosen": -0.55074441, "eval_logits/rejected": -0.51498008, "eval_logits/chosen": -0.83470297, "eval_nll_loss": 0.55109447, "eval_log_odds_ratio": -1.0913558, "eval_log_odds_chosen": 0.03414284, "epoch": 0.0234192, "global_step/max_steps": "10/1281", "percentage": "0.78%", "elapsed_time": "44m 50s", "remaining_time": "3d 23h 0m 17s"}
 {"loss": 0.67330933, "grad_norm": 0.26671985, "learning_rate": 4.26e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.003801, "rewards/chosen": -0.06494141, "rewards/rejected": -0.04647827, "rewards/accuracies": 0.0, "rewards/margins": -0.01846313, "logps/rejected": -0.46459961, "logps/chosen": -0.64990234, "logits/rejected": -0.72802734, "logits/chosen": -0.83447266, "nll_loss": 0.65087891, "log_odds_ratio": -0.93164062, "log_odds_chosen": -0.41748047, "epoch": 0.02576112, "global_step/max_steps": "11/1281", "percentage": "0.86%", "elapsed_time": "47m 58s", "remaining_time": "3d 20h 18m 12s"}
+{"loss": 0.65026855, "grad_norm": 0.19848275, "learning_rate": 4.65e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.003792, "rewards/chosen": -0.06091309, "rewards/rejected": -0.0242157, "rewards/accuracies": 0.25, "rewards/margins": -0.03674316, "logps/rejected": -0.24243164, "logps/chosen": -0.60888672, "logits/rejected": -0.48754883, "logits/chosen": -1.01074219, "nll_loss": 0.61035156, "log_odds_ratio": -1.46289062, "log_odds_chosen": -1.09619141, "epoch": 0.02810304, "global_step/max_steps": "12/1281", "percentage": "0.94%", "elapsed_time": "52m 28s", "remaining_time": "3d 20h 29m 49s"}
+{"loss": 0.71780396, "grad_norm": 0.20173968, "learning_rate": 5.04e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.003937, "rewards/chosen": -0.04553223, "rewards/rejected": -0.17755127, "rewards/accuracies": 0.25, "rewards/margins": 0.13171387, "logps/rejected": -1.7734375, "logps/chosen": -0.45458984, "logits/rejected": -0.58300781, "logits/chosen": -0.84033203, "nll_loss": 0.45361328, "log_odds_ratio": -0.74511719, "log_odds_chosen": 1.41064453, "epoch": 0.03044496, "global_step/max_steps": "13/1281", "percentage": "1.01%", "elapsed_time": "54m 46s", "remaining_time": "3d 17h 2m 59s"}
+{"loss": 0.58898926, "grad_norm": 0.16854084, "learning_rate": 5.43e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.003961, "rewards/chosen": -0.04260254, "rewards/rejected": -0.15994263, "rewards/accuracies": 0.25, "rewards/margins": 0.11721802, "logps/rejected": -1.59960938, "logps/chosen": -0.42578125, "logits/rejected": -0.49023438, "logits/chosen": -0.79003906, "nll_loss": 0.42529297, "log_odds_ratio": -0.72753906, "log_odds_chosen": 1.22216797, "epoch": 0.03278689, "global_step/max_steps": "14/1281", "percentage": "1.09%", "elapsed_time": "58m 38s", "remaining_time": "3d 16h 27m 24s"}
+{"loss": 0.58551025, "grad_norm": 0.26222301, "learning_rate": 5.81e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004044, "rewards/chosen": -0.05529785, "rewards/rejected": -0.03878784, "rewards/accuracies": 0.0, "rewards/margins": -0.01651001, "logps/rejected": -0.38818359, "logps/chosen": -0.55273438, "logits/rejected": -0.87890625, "logits/chosen": -1.02050781, "nll_loss": 0.55224609, "log_odds_ratio": -0.96386719, "log_odds_chosen": -0.47998047, "epoch": 0.03512881, "global_step/max_steps": "15/1281", "percentage": "1.17%", "elapsed_time": "1h 1m 33s", "remaining_time": "3d 14h 36m 3s"}
+{"loss": 0.605896, "grad_norm": 0.18836583, "learning_rate": 6.2e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004168, "rewards/chosen": -0.04351807, "rewards/rejected": -0.34503174, "rewards/accuracies": 0.25, "rewards/margins": 0.30126953, "logps/rejected": -3.44262695, "logps/chosen": -0.43505859, "logits/rejected": -0.60986328, "logits/chosen": -0.69921875, "nll_loss": 0.43603516, "log_odds_ratio": -0.84667969, "log_odds_chosen": 3.20458984, "epoch": 0.03747073, "global_step/max_steps": "16/1281", "percentage": "1.25%", "elapsed_time": "1h 3m 42s", "remaining_time": "3d 11h 57m 34s"}
+{"loss": 0.58584595, "grad_norm": 0.30486479, "learning_rate": 6.59e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004311, "rewards/chosen": -0.04141235, "rewards/rejected": -0.02941895, "rewards/accuracies": 0.25, "rewards/margins": -0.01199341, "logps/rejected": -0.29394531, "logps/chosen": -0.4140625, "logits/rejected": -0.50976562, "logits/chosen": -0.76806641, "nll_loss": 0.4140625, "log_odds_ratio": -0.91601562, "log_odds_chosen": -0.38378906, "epoch": 0.03981265, "global_step/max_steps": "17/1281", "percentage": "1.33%", "elapsed_time": "1h 5m 27s", "remaining_time": "3d 9h 7m 13s"}
+{"loss": 0.59320068, "grad_norm": 0.14169565, "learning_rate": 6.98e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004392, "rewards/chosen": -0.04220581, "rewards/rejected": -0.02767944, "rewards/accuracies": 0.25, "rewards/margins": -0.01452637, "logps/rejected": -0.27709961, "logps/chosen": -0.42236328, "logits/rejected": -0.61816406, "logits/chosen": -0.64355469, "nll_loss": 0.42382812, "log_odds_ratio": -0.98242188, "log_odds_chosen": -0.48632812, "epoch": 0.04215457, "global_step/max_steps": "18/1281", "percentage": "1.41%", "elapsed_time": "1h 8m 2s", "remaining_time": "3d 7h 34m 25s"}
+{"loss": 0.65429688, "grad_norm": 0.16388831, "learning_rate": 7.36e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004458, "rewards/chosen": -0.06011963, "rewards/rejected": -0.03395081, "rewards/accuracies": 0.25, "rewards/margins": -0.02618408, "logps/rejected": -0.33911133, "logps/chosen": -0.6015625, "logits/rejected": -0.35339355, "logits/chosen": -0.93847656, "nll_loss": 0.60205078, "log_odds_ratio": -1.26318359, "log_odds_chosen": -0.73242188, "epoch": 0.04449649, "global_step/max_steps": "19/1281", "percentage": "1.48%", "elapsed_time": "1h 10m 46s", "remaining_time": "3d 6h 21m 10s"}
+{"loss": 0.59362793, "grad_norm": 0.58357763, "learning_rate": 7.75e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004521, "rewards/chosen": -0.04736328, "rewards/rejected": -0.25317383, "rewards/accuracies": 0.25, "rewards/margins": 0.20574951, "logps/rejected": -2.53588867, "logps/chosen": -0.47363281, "logits/rejected": -0.48962402, "logits/chosen": -0.78320312, "nll_loss": 0.47216797, "log_odds_ratio": -0.85839844, "log_odds_chosen": 2.02490234, "epoch": 0.04683841, "global_step/max_steps": "20/1281", "percentage": "1.56%", "elapsed_time": "1h 13m 28s", "remaining_time": "3d 5h 12m 45s"}
+{"eval_loss": 0.60351562, "eval_runtime": 641.8534, "eval_samples_per_second": 0.657, "eval_steps_per_second": 0.083, "eval_rewards/chosen": -0.05049335, "eval_rewards/rejected": -0.07673041, "eval_rewards/accuracies": 0.1509434, "eval_rewards/margins": 0.02624972, "eval_logps/rejected": -0.76708984, "eval_logps/chosen": -0.50484598, "eval_logits/rejected": -0.43351543, "eval_logits/chosen": -0.73570168, "eval_nll_loss": 0.50490123, "eval_log_odds_ratio": -1.04105711, "eval_log_odds_chosen": -0.06548496, "epoch": 0.04683841, "global_step/max_steps": "20/1281", "percentage": "1.56%", "elapsed_time": "1h 24m 10s", "remaining_time": "3d 16h 27m 14s"}
+{"loss": 0.57366943, "grad_norm": 0.14497812, "learning_rate": 8.14e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004005, "rewards/chosen": -0.0569458, "rewards/rejected": -0.03063965, "rewards/accuracies": 0.25, "rewards/margins": -0.02630615, "logps/rejected": -0.30639648, "logps/chosen": -0.56933594, "logits/rejected": -0.45800781, "logits/chosen": -0.69482422, "nll_loss": 0.5703125, "log_odds_ratio": -1.23632812, "log_odds_chosen": -0.80859375, "epoch": 0.04918033, "global_step/max_steps": "21/1281", "percentage": "1.64%", "elapsed_time": "1h 27m 7s", "remaining_time": "3d 15h 7m 51s"}