Training in progress, step 120

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6c26e16b214aa881cd57aa81df76ab502fe19c6e2cccce10b24ab034b343c43b
+oid sha256:f25463bb5f45f6f9bc1c30182e3e84c708fa721d54179e8a412b2f959349fdf6
 size 6627156248

logging.jsonl

@@ -119,3 +119,14 @@
 {"loss": 0.39370728, "grad_norm": 0.13373983, "learning_rate": 4.842e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003328, "rewards/chosen": -0.03863525, "rewards/rejected": -0.08886719, "rewards/accuracies": 0.75, "rewards/margins": 0.0501709, "logps/rejected": -0.88818359, "logps/chosen": -0.38623047, "logits/rejected": -0.45410156, "logits/chosen": -0.28271484, "nll_loss": 0.38623047, "log_odds_ratio": -0.53930664, "log_odds_chosen": 0.86914062, "epoch": 0.38652482, "global_step/max_steps": "109/846", "percentage": "12.88%", "elapsed_time": "9h 5m 32s", "remaining_time": "2d 13h 28m 42s"}
 {"loss": 0.38543701, "grad_norm": 0.35314423, "learning_rate": 4.836e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003337, "rewards/chosen": -0.03192139, "rewards/rejected": -0.0549469, "rewards/accuracies": 0.5, "rewards/margins": 0.02294922, "logps/rejected": -0.54992676, "logps/chosen": -0.31933594, "logits/rejected": -0.18151855, "logits/chosen": -0.5369873, "nll_loss": 0.3190918, "log_odds_ratio": -0.69104004, "log_odds_chosen": 0.38671875, "epoch": 0.39007092, "global_step/max_steps": "110/846", "percentage": "13.00%", "elapsed_time": "9h 9m 6s", "remaining_time": "2d 13h 14m 0s"}
 {"eval_loss": 0.42397514, "eval_runtime": 568.2834, "eval_samples_per_second": 0.491, "eval_steps_per_second": 0.062, "eval_rewards/chosen": -0.04015939, "eval_rewards/rejected": -0.09039132, "eval_rewards/accuracies": 0.5714286, "eval_rewards/margins": 0.05023368, "eval_logps/rejected": -0.90404576, "eval_logps/chosen": -0.40133929, "eval_logits/rejected": -0.27489188, "eval_logits/chosen": -0.40006277, "eval_nll_loss": 0.40170202, "eval_log_odds_ratio": -0.54534388, "eval_log_odds_chosen": 0.8895089, "epoch": 0.39007092, "global_step/max_steps": "110/846", "percentage": "13.00%", "elapsed_time": "9h 18m 34s", "remaining_time": "2d 14h 17m 22s"}
+{"loss": 0.40066528, "grad_norm": 0.31577688, "learning_rate": 4.829e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003276, "rewards/chosen": -0.03585815, "rewards/rejected": -0.07455444, "rewards/accuracies": 0.5, "rewards/margins": 0.03875732, "logps/rejected": -0.74389648, "logps/chosen": -0.35864258, "logits/rejected": -0.28894043, "logits/chosen": -0.38806152, "nll_loss": 0.35791016, "log_odds_ratio": -0.58721924, "log_odds_chosen": 0.69628906, "epoch": 0.39361702, "global_step/max_steps": "111/846", "percentage": "13.12%", "elapsed_time": "9h 24m 28s", "remaining_time": "2d 14h 17m 45s"}
+{"loss": 0.43609619, "grad_norm": 0.14331454, "learning_rate": 4.823e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003273, "rewards/chosen": -0.02960205, "rewards/rejected": -0.13140869, "rewards/accuracies": 1.0, "rewards/margins": 0.10192871, "logps/rejected": -1.31494141, "logps/chosen": -0.29614258, "logits/rejected": -0.2755127, "logits/chosen": -0.26312256, "nll_loss": 0.29589844, "log_odds_ratio": -0.23156738, "log_odds_chosen": 1.83959961, "epoch": 0.39716312, "global_step/max_steps": "112/846", "percentage": "13.24%", "elapsed_time": "9h 30m 1s", "remaining_time": "2d 14h 15m 39s"}
+{"loss": 0.39654541, "grad_norm": 0.21657935, "learning_rate": 4.816e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003285, "rewards/chosen": -0.03845215, "rewards/rejected": -0.07147217, "rewards/accuracies": 0.5, "rewards/margins": 0.03308105, "logps/rejected": -0.71533203, "logps/chosen": -0.38427734, "logits/rejected": -0.51269531, "logits/chosen": -0.60058594, "nll_loss": 0.38378906, "log_odds_ratio": -0.66967773, "log_odds_chosen": 0.41503906, "epoch": 0.40070922, "global_step/max_steps": "113/846", "percentage": "13.36%", "elapsed_time": "9h 33m 3s", "remaining_time": "2d 13h 57m 13s"}
+{"loss": 0.43988037, "grad_norm": 0.17497085, "learning_rate": 4.809e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.00329, "rewards/chosen": -0.02536011, "rewards/rejected": -0.05245972, "rewards/accuracies": 1.0, "rewards/margins": 0.02713013, "logps/rejected": -0.52441406, "logps/chosen": -0.25341797, "logits/rejected": -0.56201172, "logits/chosen": -0.37255859, "nll_loss": 0.25317383, "log_odds_ratio": -0.37939453, "log_odds_chosen": 0.82861328, "epoch": 0.40425532, "global_step/max_steps": "114/846", "percentage": "13.48%", "elapsed_time": "9h 37m 15s", "remaining_time": "2d 13h 46m 38s"}
+{"loss": 0.4364624, "grad_norm": 0.19731806, "learning_rate": 4.803e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003285, "rewards/chosen": -0.03411865, "rewards/rejected": -0.05270386, "rewards/accuracies": 0.75, "rewards/margins": 0.01858521, "logps/rejected": -0.52587891, "logps/chosen": -0.34179688, "logits/rejected": -0.28845215, "logits/chosen": -0.24676514, "nll_loss": 0.34228516, "log_odds_ratio": -0.55419922, "log_odds_chosen": 0.50439453, "epoch": 0.40780142, "global_step/max_steps": "115/846", "percentage": "13.59%", "elapsed_time": "9h 43m 6s", "remaining_time": "2d 13h 46m 34s"}
+{"loss": 0.36676025, "grad_norm": 0.16225766, "learning_rate": 4.796e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003293, "rewards/chosen": -0.02584839, "rewards/rejected": -0.15896606, "rewards/accuracies": 0.75, "rewards/margins": 0.13305664, "logps/rejected": -1.58740234, "logps/chosen": -0.25830078, "logits/rejected": -0.32080078, "logits/chosen": -0.29064941, "nll_loss": 0.25805664, "log_odds_ratio": -0.29370117, "log_odds_chosen": 2.08496094, "epoch": 0.41134752, "global_step/max_steps": "116/846", "percentage": "13.71%", "elapsed_time": "9h 46m 45s", "remaining_time": "2d 13h 32m 34s"}
+{"loss": 0.37265015, "grad_norm": 0.13713335, "learning_rate": 4.79e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.0033, "rewards/chosen": -0.03543091, "rewards/rejected": -0.04107666, "rewards/accuracies": 0.5, "rewards/margins": 0.00564575, "logps/rejected": -0.41088867, "logps/chosen": -0.35400391, "logits/rejected": -0.50610352, "logits/chosen": -0.66552734, "nll_loss": 0.35400391, "log_odds_ratio": -0.7277832, "log_odds_chosen": 0.10253906, "epoch": 0.41489362, "global_step/max_steps": "117/846", "percentage": "13.83%", "elapsed_time": "9h 50m 38s", "remaining_time": "2d 13h 20m 9s"}
+{"loss": 0.39312744, "grad_norm": 0.14967194, "learning_rate": 4.783e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003307, "rewards/chosen": -0.0256958, "rewards/rejected": -0.08746338, "rewards/accuracies": 0.75, "rewards/margins": 0.06167603, "logps/rejected": -0.87597656, "logps/chosen": -0.25708008, "logits/rejected": -0.46923828, "logits/chosen": -0.31445312, "nll_loss": 0.25634766, "log_odds_ratio": -0.41497803, "log_odds_chosen": 1.15649414, "epoch": 0.41843972, "global_step/max_steps": "118/846", "percentage": "13.95%", "elapsed_time": "9h 54m 25s", "remaining_time": "2d 13h 7m 15s"}
+{"loss": 0.37322998, "grad_norm": 0.11315849, "learning_rate": 4.777e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003303, "rewards/chosen": -0.02914429, "rewards/rejected": -0.06541443, "rewards/accuracies": 0.25, "rewards/margins": 0.03622437, "logps/rejected": -0.65368652, "logps/chosen": -0.29174805, "logits/rejected": -0.4230957, "logits/chosen": -0.60546875, "nll_loss": 0.29223633, "log_odds_ratio": -0.75622559, "log_odds_chosen": 0.35253906, "epoch": 0.42198582, "global_step/max_steps": "119/846", "percentage": "14.07%", "elapsed_time": "10h 0m 6s", "remaining_time": "2d 13h 6m 14s"}
+{"loss": 0.37789917, "grad_norm": 0.1166358, "learning_rate": 4.77e-05, "memory(GiB)": 132.99, "train_speed(iter/s)": 0.003305, "rewards/chosen": -0.04443359, "rewards/rejected": -0.11999512, "rewards/accuracies": 1.0, "rewards/margins": 0.07550049, "logps/rejected": -1.20019531, "logps/chosen": -0.4440918, "logits/rejected": -0.24209595, "logits/chosen": -0.18579102, "nll_loss": 0.4440918, "log_odds_ratio": -0.31967163, "log_odds_chosen": 1.44921875, "epoch": 0.42553191, "global_step/max_steps": "120/846", "percentage": "14.18%", "elapsed_time": "10h 4m 53s", "remaining_time": "2d 12h 59m 36s"}
+{"eval_loss": 0.41283742, "eval_runtime": 568.1176, "eval_samples_per_second": 0.491, "eval_steps_per_second": 0.062, "eval_rewards/chosen": -0.03860386, "eval_rewards/rejected": -0.12836914, "eval_rewards/accuracies": 0.65714288, "eval_rewards/margins": 0.08977748, "eval_logps/rejected": -1.28404021, "eval_logps/chosen": -0.38635603, "eval_logits/rejected": -0.39749581, "eval_logits/chosen": -0.51142579, "eval_nll_loss": 0.38696986, "eval_log_odds_ratio": -0.48230678, "eval_log_odds_chosen": 1.40002787, "epoch": 0.42553191, "global_step/max_steps": "120/846", "percentage": "14.18%", "elapsed_time": "10h 14m 21s", "remaining_time": "2d 13h 56m 53s"}