Training in progress, step 50

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ada23fabdae7fdde487b7d5345db0f0adfb3bceb97e236ad2223ca5ced5caf4a
+oid sha256:741224a7698ae2da8e0df9e2b256d7749de879b16b47c34cee5c9e628b743609
 size 6627156248

logging.jsonl

@@ -43,3 +43,14 @@
 {"loss": 0.50769043, "grad_norm": 0.14049903, "learning_rate": 1.55e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004395, "rewards/chosen": -0.05560303, "rewards/rejected": -0.027771, "rewards/accuracies": 0.25, "rewards/margins": -0.02780151, "logps/rejected": -0.27783203, "logps/chosen": -0.55664062, "logits/rejected": -0.26611328, "logits/chosen": -0.61376953, "nll_loss": 0.5546875, "log_odds_ratio": -1.17773438, "log_odds_chosen": -0.75634766, "epoch": 0.09367681, "global_step/max_steps": "40/1281", "percentage": "3.12%", "elapsed_time": "2h 31m 25s", "remaining_time": "3d 6h 17m 58s"}
 {"eval_loss": 0.54541248, "eval_runtime": 642.0806, "eval_samples_per_second": 0.657, "eval_steps_per_second": 0.083, "eval_rewards/chosen": -0.04489654, "eval_rewards/rejected": -0.09570197, "eval_rewards/accuracies": 0.16981132, "eval_rewards/margins": 0.05081004, "eval_logps/rejected": -0.95734447, "eval_logps/chosen": -0.44903448, "eval_logits/rejected": -0.26981685, "eval_logits/chosen": -0.51291418, "eval_nll_loss": 0.44921875, "eval_log_odds_ratio": -0.98677033, "eval_log_odds_chosen": 0.21937647, "epoch": 0.09367681, "global_step/max_steps": "40/1281", "percentage": "3.12%", "elapsed_time": "2h 42m 7s", "remaining_time": "3d 11h 49m 58s"}
 {"loss": 0.55899048, "grad_norm": 0.17105363, "learning_rate": 1.589e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004123, "rewards/chosen": -0.03289795, "rewards/rejected": -1.04510498, "rewards/accuracies": 0.25, "rewards/margins": 1.01303101, "logps/rejected": -10.45092773, "logps/chosen": -0.32885742, "logits/rejected": -0.27050781, "logits/chosen": -0.44433594, "nll_loss": 0.32739258, "log_odds_ratio": -0.75488281, "log_odds_chosen": 10.17626953, "epoch": 0.09601874, "global_step/max_steps": "41/1281", "percentage": "3.20%", "elapsed_time": "2h 45m 28s", "remaining_time": "3d 11h 24m 50s"}
+{"loss": 0.55685425, "grad_norm": 0.19478773, "learning_rate": 1.628e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004155, "rewards/chosen": -0.04776001, "rewards/rejected": -0.45196533, "rewards/accuracies": 0.25, "rewards/margins": 0.40423584, "logps/rejected": -4.51586914, "logps/chosen": -0.47827148, "logits/rejected": -0.33740234, "logits/chosen": -0.515625, "nll_loss": 0.47875977, "log_odds_ratio": -0.91015625, "log_odds_chosen": 3.99169922, "epoch": 0.09836066, "global_step/max_steps": "42/1281", "percentage": "3.28%", "elapsed_time": "2h 48m 13s", "remaining_time": "3d 10h 42m 33s"}
+{"loss": 0.47839355, "grad_norm": 0.12849084, "learning_rate": 1.667e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.00416, "rewards/chosen": -0.04858398, "rewards/rejected": -0.95053101, "rewards/accuracies": 0.25, "rewards/margins": 0.90255737, "logps/rejected": -9.49780273, "logps/chosen": -0.48583984, "logits/rejected": -0.46972656, "logits/chosen": -0.58349609, "nll_loss": 0.48535156, "log_odds_ratio": -0.7890625, "log_odds_chosen": 9.04199219, "epoch": 0.10070258, "global_step/max_steps": "43/1281", "percentage": "3.36%", "elapsed_time": "2h 52m 1s", "remaining_time": "3d 10h 32m 38s"}
+{"loss": 0.43128967, "grad_norm": 0.12454039, "learning_rate": 1.705e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004207, "rewards/chosen": -0.02954102, "rewards/rejected": -0.02560425, "rewards/accuracies": 0.0, "rewards/margins": -0.00393677, "logps/rejected": -0.25610352, "logps/chosen": -0.29492188, "logits/rejected": -0.37597656, "logits/chosen": -0.38818359, "nll_loss": 0.29492188, "log_odds_ratio": -0.7734375, "log_odds_chosen": -0.15405273, "epoch": 0.1030445, "global_step/max_steps": "44/1281", "percentage": "3.43%", "elapsed_time": "2h 54m 3s", "remaining_time": "3d 9h 33m 26s"}
+{"loss": 0.49902344, "grad_norm": 0.14915697, "learning_rate": 1.744e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004261, "rewards/chosen": -0.03634644, "rewards/rejected": -1.38952637, "rewards/accuracies": 0.5, "rewards/margins": 1.35101318, "logps/rejected": -13.86376953, "logps/chosen": -0.36328125, "logits/rejected": -0.28930664, "logits/chosen": -0.43432617, "nll_loss": 0.36376953, "log_odds_ratio": -0.49902344, "log_odds_chosen": 14.046875, "epoch": 0.10538642, "global_step/max_steps": "45/1281", "percentage": "3.51%", "elapsed_time": "2h 55m 44s", "remaining_time": "3d 8h 27m 3s"}
+{"loss": 0.46038818, "grad_norm": 0.12378819, "learning_rate": 1.783e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004317, "rewards/chosen": -0.03189087, "rewards/rejected": -0.03677368, "rewards/accuracies": 0.25, "rewards/margins": 0.00491333, "logps/rejected": -0.3684082, "logps/chosen": -0.31884766, "logits/rejected": -0.34179688, "logits/chosen": -0.42041016, "nll_loss": 0.31835938, "log_odds_ratio": -0.72192383, "log_odds_chosen": 0.12817383, "epoch": 0.10772834, "global_step/max_steps": "46/1281", "percentage": "3.59%", "elapsed_time": "2h 57m 20s", "remaining_time": "3d 7h 21m 1s"}
+{"loss": 0.4954834, "grad_norm": 0.67944998, "learning_rate": 1.822e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004311, "rewards/chosen": -0.03808594, "rewards/rejected": -0.46389771, "rewards/accuracies": 0.5, "rewards/margins": 0.42666626, "logps/rejected": -4.64770508, "logps/chosen": -0.38110352, "logits/rejected": -0.31884766, "logits/chosen": -0.40136719, "nll_loss": 0.38134766, "log_odds_ratio": -0.62208557, "log_odds_chosen": 4.67871094, "epoch": 0.11007026, "global_step/max_steps": "47/1281", "percentage": "3.67%", "elapsed_time": "3h 1m 27s", "remaining_time": "3d 7h 24m 25s"}
+{"loss": 0.48057556, "grad_norm": 0.11546308, "learning_rate": 1.86e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004323, "rewards/chosen": -0.04058838, "rewards/rejected": -0.31646729, "rewards/accuracies": 0.5, "rewards/margins": 0.27679443, "logps/rejected": -3.16845703, "logps/chosen": -0.40625, "logits/rejected": -0.46679688, "logits/chosen": -0.50732422, "nll_loss": 0.40625, "log_odds_ratio": -0.63623047, "log_odds_chosen": 3.00048828, "epoch": 0.11241218, "global_step/max_steps": "48/1281", "percentage": "3.75%", "elapsed_time": "3h 4m 47s", "remaining_time": "3d 7h 6m 49s"}
+{"loss": 0.60256958, "grad_norm": 0.73745817, "learning_rate": 1.899e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004338, "rewards/chosen": -0.03192139, "rewards/rejected": -0.02645874, "rewards/accuracies": 0.25, "rewards/margins": -0.00546265, "logps/rejected": -0.26489258, "logps/chosen": -0.31884766, "logits/rejected": -0.28442383, "logits/chosen": -0.30834961, "nll_loss": 0.31787109, "log_odds_ratio": -0.79199219, "log_odds_chosen": -0.17993164, "epoch": 0.1147541, "global_step/max_steps": "49/1281", "percentage": "3.83%", "elapsed_time": "3h 8m 0s", "remaining_time": "3d 6h 47m 0s"}
+{"loss": 0.60345459, "grad_norm": 1.77461624, "learning_rate": 1.938e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004356, "rewards/chosen": -0.03149414, "rewards/rejected": -0.02838135, "rewards/accuracies": 0.25, "rewards/margins": -0.00311279, "logps/rejected": -0.28344727, "logps/chosen": -0.31469727, "logits/rejected": -0.31982422, "logits/chosen": -0.47070312, "nll_loss": 0.31445312, "log_odds_ratio": -0.76757812, "log_odds_chosen": -0.14038086, "epoch": 0.11709602, "global_step/max_steps": "50/1281", "percentage": "3.90%", "elapsed_time": "3h 11m 2s", "remaining_time": "3d 6h 23m 16s"}
+{"eval_loss": 0.52925986, "eval_runtime": 641.4732, "eval_samples_per_second": 0.658, "eval_steps_per_second": 0.083, "eval_rewards/chosen": -0.04372421, "eval_rewards/rejected": -0.08673556, "eval_rewards/accuracies": 0.16981132, "eval_rewards/margins": 0.04298257, "eval_logps/rejected": -0.86737174, "eval_logps/chosen": -0.43724203, "eval_logits/rejected": -0.28344497, "eval_logits/chosen": -0.52488852, "eval_nll_loss": 0.43731573, "eval_log_odds_ratio": -0.95137572, "eval_log_odds_chosen": 0.18302256, "epoch": 0.11709602, "global_step/max_steps": "50/1281", "percentage": "3.90%", "elapsed_time": "3h 21m 43s", "remaining_time": "3d 10h 46m 29s"}
+{"loss": 0.52334595, "grad_norm": 0.1435816, "learning_rate": 1.977e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004131, "rewards/chosen": -0.03616333, "rewards/rejected": -0.08886719, "rewards/accuracies": 0.25, "rewards/margins": 0.05279541, "logps/rejected": -0.8894043, "logps/chosen": -0.36132812, "logits/rejected": -0.44628906, "logits/chosen": -0.41162109, "nll_loss": 0.36132812, "log_odds_ratio": -0.69046021, "log_odds_chosen": 0.76855469, "epoch": 0.11943794, "global_step/max_steps": "51/1281", "percentage": "3.98%", "elapsed_time": "3h 25m 29s", "remaining_time": "3d 10h 36m 1s"}