Training in progress, step 60

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:741224a7698ae2da8e0df9e2b256d7749de879b16b47c34cee5c9e628b743609
+oid sha256:eef43ce4316674099e870be849acb3a0a06361061c542f2dca84ae139a16c51e
 size 6627156248

logging.jsonl

@@ -54,3 +54,14 @@
 {"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"}
+{"loss": 0.56259155, "grad_norm": 0.14094751, "learning_rate": 2.016e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004126, "rewards/chosen": -0.046875, "rewards/rejected": -0.03430176, "rewards/accuracies": 0.25, "rewards/margins": -0.01257324, "logps/rejected": -0.34326172, "logps/chosen": -0.46923828, "logits/rejected": -0.46655273, "logits/chosen": -0.51611328, "nll_loss": 0.46923828, "log_odds_ratio": -0.92871094, "log_odds_chosen": -0.41503906, "epoch": 0.12177986, "global_step/max_steps": "52/1281", "percentage": "4.06%", "elapsed_time": "3h 29m 47s", "remaining_time": "3d 10h 38m 28s"}
+{"loss": 0.52056885, "grad_norm": 0.15958506, "learning_rate": 2.054e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004139, "rewards/chosen": -0.04098511, "rewards/rejected": -0.02426147, "rewards/accuracies": 0.25, "rewards/margins": -0.01672363, "logps/rejected": -0.24267578, "logps/chosen": -0.40966797, "logits/rejected": -0.22973633, "logits/chosen": -0.54760742, "nll_loss": 0.40917969, "log_odds_ratio": -1.02636719, "log_odds_chosen": -0.52807617, "epoch": 0.12412178, "global_step/max_steps": "53/1281", "percentage": "4.14%", "elapsed_time": "3h 33m 9s", "remaining_time": "3d 10h 18m 56s"}
+{"loss": 0.51293945, "grad_norm": 0.13775666, "learning_rate": 2.093e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004177, "rewards/chosen": -0.05709839, "rewards/rejected": -0.02593994, "rewards/accuracies": 0.25, "rewards/margins": -0.03112793, "logps/rejected": -0.25952148, "logps/chosen": -0.57080078, "logits/rejected": -0.24047852, "logits/chosen": -0.68408203, "nll_loss": 0.57177734, "log_odds_ratio": -1.27734375, "log_odds_chosen": -0.89746094, "epoch": 0.1264637, "global_step/max_steps": "54/1281", "percentage": "4.22%", "elapsed_time": "3h 35m 11s", "remaining_time": "3d 9h 29m 37s"}
+{"loss": 0.50646973, "grad_norm": 0.16116208, "learning_rate": 2.132e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.00421, "rewards/chosen": -0.03811646, "rewards/rejected": -0.0274353, "rewards/accuracies": 0.25, "rewards/margins": -0.01068115, "logps/rejected": -0.27490234, "logps/chosen": -0.38134766, "logits/rejected": -0.34521484, "logits/chosen": -0.52001953, "nll_loss": 0.38085938, "log_odds_ratio": -0.95605469, "log_odds_chosen": -0.44189453, "epoch": 0.12880562, "global_step/max_steps": "55/1281", "percentage": "4.29%", "elapsed_time": "3h 37m 29s", "remaining_time": "3d 8h 48m 9s"}
+{"loss": 0.52423096, "grad_norm": 0.27805322, "learning_rate": 2.171e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.00425, "rewards/chosen": -0.06137085, "rewards/rejected": -0.02838135, "rewards/accuracies": 0.0, "rewards/margins": -0.0329895, "logps/rejected": -0.28393555, "logps/chosen": -0.61425781, "logits/rejected": -0.24029541, "logits/chosen": -0.66894531, "nll_loss": 0.61425781, "log_odds_ratio": -1.24023438, "log_odds_chosen": -0.86279297, "epoch": 0.13114754, "global_step/max_steps": "56/1281", "percentage": "4.37%", "elapsed_time": "3h 39m 20s", "remaining_time": "3d 7h 58m 11s"}
+{"loss": 0.58349609, "grad_norm": 0.37276149, "learning_rate": 2.209e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004256, "rewards/chosen": -0.03268433, "rewards/rejected": -0.03115845, "rewards/accuracies": 0.5, "rewards/margins": -0.00152588, "logps/rejected": -0.31176758, "logps/chosen": -0.32666016, "logits/rejected": -0.22851562, "logits/chosen": -0.66503906, "nll_loss": 0.32641602, "log_odds_ratio": -0.72753906, "log_odds_chosen": -0.05810547, "epoch": 0.13348946, "global_step/max_steps": "57/1281", "percentage": "4.45%", "elapsed_time": "3h 42m 56s", "remaining_time": "3d 7h 47m 31s"}
+{"loss": 0.5043335, "grad_norm": 0.12889104, "learning_rate": 2.248e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004256, "rewards/chosen": -0.03897095, "rewards/rejected": -0.33685303, "rewards/accuracies": 0.25, "rewards/margins": 0.29818726, "logps/rejected": -3.36035156, "logps/chosen": -0.38964844, "logits/rejected": -0.18151855, "logits/chosen": -0.54736328, "nll_loss": 0.39013672, "log_odds_ratio": -0.78027344, "log_odds_chosen": 3.05664062, "epoch": 0.13583138, "global_step/max_steps": "58/1281", "percentage": "4.53%", "elapsed_time": "3h 46m 51s", "remaining_time": "3d 7h 43m 40s"}
+{"loss": 0.47988892, "grad_norm": 0.13205451, "learning_rate": 2.287e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004261, "rewards/chosen": -0.03903198, "rewards/rejected": -0.04699707, "rewards/accuracies": 0.5, "rewards/margins": 0.00796509, "logps/rejected": -0.47021484, "logps/chosen": -0.39013672, "logits/rejected": -0.52685547, "logits/chosen": -0.64306641, "nll_loss": 0.38916016, "log_odds_ratio": -0.70678711, "log_odds_chosen": 0.13916016, "epoch": 0.1381733, "global_step/max_steps": "59/1281", "percentage": "4.61%", "elapsed_time": "3h 50m 31s", "remaining_time": "3d 7h 34m 45s"}
+{"loss": 0.54541016, "grad_norm": 0.1370042, "learning_rate": 2.326e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004286, "rewards/chosen": -0.05780029, "rewards/rejected": -0.03494263, "rewards/accuracies": 0.25, "rewards/margins": -0.02288818, "logps/rejected": -0.34912109, "logps/chosen": -0.578125, "logits/rejected": -0.58300781, "logits/chosen": -0.58691406, "nll_loss": 0.57763672, "log_odds_ratio": -1.07226562, "log_odds_chosen": -0.60742188, "epoch": 0.14051522, "global_step/max_steps": "60/1281", "percentage": "4.68%", "elapsed_time": "3h 53m 2s", "remaining_time": "3d 7h 2m 17s"}
+{"eval_loss": 0.51840079, "eval_runtime": 642.1861, "eval_samples_per_second": 0.657, "eval_steps_per_second": 0.083, "eval_rewards/chosen": -0.04324974, "eval_rewards/rejected": -0.06017721, "eval_rewards/accuracies": 0.26415095, "eval_rewards/margins": 0.01688255, "eval_logps/rejected": -0.60198629, "eval_logps/chosen": -0.43261719, "eval_logits/rejected": -0.27153504, "eval_logits/chosen": -0.53526682, "eval_nll_loss": 0.43278301, "eval_log_odds_ratio": -0.87953389, "eval_log_odds_chosen": 0.00254275, "epoch": 0.14051522, "global_step/max_steps": "60/1281", "percentage": "4.68%", "elapsed_time": "4h 3m 44s", "remaining_time": "3d 10h 40m 6s"}
+{"loss": 0.52972412, "grad_norm": 0.14679687, "learning_rate": 2.364e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004106, "rewards/chosen": -0.04101562, "rewards/rejected": -0.03057861, "rewards/accuracies": 0.25, "rewards/margins": -0.01043701, "logps/rejected": -0.30566406, "logps/chosen": -0.41064453, "logits/rejected": -0.15844727, "logits/chosen": -0.48388672, "nll_loss": 0.40869141, "log_odds_ratio": -0.86425781, "log_odds_chosen": -0.26953125, "epoch": 0.14285714, "global_step/max_steps": "61/1281", "percentage": "4.76%", "elapsed_time": "4h 7m 20s", "remaining_time": "3d 10h 26m 53s"}