Training in progress, step 170

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:39aa253082dd7775361bcc145ce2e3d84766cf5c98c02e5b9a69643f16b4f492
+oid sha256:c9e1e53aa81d6d2fd33a8c09a3191c0c5d8dc036e5661849d30ee6590433d2dd
 size 6627156248

logging.jsonl

@@ -174,3 +174,14 @@
 {"loss": 0.53839111, "grad_norm": 0.12544316, "learning_rate": 3.175e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.004023, "rewards/chosen": -0.07232666, "rewards/rejected": -0.04547119, "rewards/accuracies": 0.75, "rewards/margins": -0.0267334, "logps/rejected": -0.45458984, "logps/chosen": -0.72314453, "logits/rejected": -0.15310669, "logits/chosen": -1.04785156, "nll_loss": 0.72558594, "log_odds_ratio": -1.03613281, "log_odds_chosen": -0.40722656, "epoch": 0.37236534, "global_step/max_steps": "159/427", "percentage": "37.24%", "elapsed_time": "10h 58m 8s", "remaining_time": "18h 29m 19s"}
 {"loss": 0.45611572, "grad_norm": 0.09365269, "learning_rate": 3.164e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.00403, "rewards/chosen": -0.0324707, "rewards/rejected": -0.02557373, "rewards/accuracies": 0.0, "rewards/margins": -0.00689697, "logps/rejected": -0.25561523, "logps/chosen": -0.32446289, "logits/rejected": -0.53125, "logits/chosen": -0.45849609, "nll_loss": 0.32470703, "log_odds_ratio": -0.84277344, "log_odds_chosen": -0.27441406, "epoch": 0.37470726, "global_step/max_steps": "160/427", "percentage": "37.47%", "elapsed_time": "11h 1m 5s", "remaining_time": "18h 23m 12s"}
 {"eval_loss": 0.46463084, "eval_runtime": 649.616, "eval_samples_per_second": 0.65, "eval_steps_per_second": 0.082, "eval_rewards/chosen": -0.0413565, "eval_rewards/rejected": -0.12970085, "eval_rewards/accuracies": 0.60377359, "eval_rewards/margins": 0.08834436, "eval_logps/rejected": -1.29744625, "eval_logps/chosen": -0.41352814, "eval_logits/rejected": -0.28135133, "eval_logits/chosen": -0.75331199, "eval_nll_loss": 0.41358343, "eval_log_odds_ratio": -0.58239979, "eval_log_odds_chosen": 1.09166789, "epoch": 0.37470726, "global_step/max_steps": "160/427", "percentage": "37.47%", "elapsed_time": "11h 11m 55s", "remaining_time": "18h 41m 16s"}
+{"loss": 0.46447754, "grad_norm": 0.08780798, "learning_rate": 3.152e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003955, "rewards/chosen": -0.04327393, "rewards/rejected": -0.03875732, "rewards/accuracies": 0.5, "rewards/margins": -0.0045166, "logps/rejected": -0.38745117, "logps/chosen": -0.43261719, "logits/rejected": -0.60791016, "logits/chosen": -0.59570312, "nll_loss": 0.43212891, "log_odds_ratio": -0.84228516, "log_odds_chosen": -0.18798828, "epoch": 0.37704918, "global_step/max_steps": "161/427", "percentage": "37.70%", "elapsed_time": "11h 17m 49s", "remaining_time": "18h 39m 53s"}
+{"loss": 0.45516968, "grad_norm": 0.09269176, "learning_rate": 3.14e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003961, "rewards/chosen": -0.03967285, "rewards/rejected": -0.64620972, "rewards/accuracies": 0.75, "rewards/margins": 0.60678101, "logps/rejected": -6.47705078, "logps/chosen": -0.39648438, "logits/rejected": -0.44628906, "logits/chosen": -0.87988281, "nll_loss": 0.39697266, "log_odds_ratio": -0.34814453, "log_odds_chosen": 6.62597656, "epoch": 0.3793911, "global_step/max_steps": "162/427", "percentage": "37.94%", "elapsed_time": "11h 21m 7s", "remaining_time": "18h 34m 10s"}
+{"loss": 0.427948, "grad_norm": 0.07641727, "learning_rate": 3.128e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003968, "rewards/chosen": -0.02523804, "rewards/rejected": -0.20239258, "rewards/accuracies": 0.75, "rewards/margins": 0.17672729, "logps/rejected": -2.02050781, "logps/chosen": -0.25244141, "logits/rejected": -0.15368652, "logits/chosen": -0.63916016, "nll_loss": 0.25244141, "log_odds_ratio": -0.32617188, "log_odds_chosen": 2.53515625, "epoch": 0.38173302, "global_step/max_steps": "163/427", "percentage": "38.17%", "elapsed_time": "11h 24m 5s", "remaining_time": "18h 27m 58s"}
+{"loss": 0.41162109, "grad_norm": 0.07814033, "learning_rate": 3.116e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003979, "rewards/chosen": -0.03631592, "rewards/rejected": -0.05712891, "rewards/accuracies": 0.75, "rewards/margins": 0.02081299, "logps/rejected": -0.57080078, "logps/chosen": -0.36352539, "logits/rejected": -0.2756424, "logits/chosen": -0.84960938, "nll_loss": 0.36352539, "log_odds_ratio": -0.50732422, "log_odds_chosen": 0.47094727, "epoch": 0.38407494, "global_step/max_steps": "164/427", "percentage": "38.41%", "elapsed_time": "11h 26m 17s", "remaining_time": "18h 20m 34s"}
+{"loss": 0.42019653, "grad_norm": 0.08304253, "learning_rate": 3.104e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003983, "rewards/chosen": -0.04974365, "rewards/rejected": -0.04663086, "rewards/accuracies": 0.5, "rewards/margins": -0.00311279, "logps/rejected": -0.46679688, "logps/chosen": -0.49731445, "logits/rejected": -0.28320312, "logits/chosen": -1.06054688, "nll_loss": 0.49853516, "log_odds_ratio": -0.72607422, "log_odds_chosen": 0.046875, "epoch": 0.38641686, "global_step/max_steps": "165/427", "percentage": "38.64%", "elapsed_time": "11h 29m 53s", "remaining_time": "18h 15m 28s"}
+{"loss": 0.45248413, "grad_norm": 0.25102618, "learning_rate": 3.092e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003985, "rewards/chosen": -0.03695679, "rewards/rejected": -0.05917358, "rewards/accuracies": 1.0, "rewards/margins": 0.0222168, "logps/rejected": -0.59130859, "logps/chosen": -0.36962891, "logits/rejected": -0.41699219, "logits/chosen": -0.70556641, "nll_loss": 0.36938477, "log_odds_ratio": -0.50170898, "log_odds_chosen": 0.51025391, "epoch": 0.38875878, "global_step/max_steps": "166/427", "percentage": "38.88%", "elapsed_time": "11h 33m 41s", "remaining_time": "18h 10m 40s"}
+{"loss": 0.5090332, "grad_norm": 0.10502508, "learning_rate": 3.081e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.00399, "rewards/chosen": -0.04241943, "rewards/rejected": -0.04272461, "rewards/accuracies": 0.5, "rewards/margins": 0.00030518, "logps/rejected": -0.42773438, "logps/chosen": -0.42480469, "logits/rejected": -0.32324219, "logits/chosen": -0.65576172, "nll_loss": 0.42480469, "log_odds_ratio": -0.703125, "log_odds_chosen": -0.00390625, "epoch": 0.3911007, "global_step/max_steps": "167/427", "percentage": "39.11%", "elapsed_time": "11h 37m 1s", "remaining_time": "18h 5m 10s"}
+{"loss": 0.45736694, "grad_norm": 0.0924931, "learning_rate": 3.069e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.00399, "rewards/chosen": -0.05560303, "rewards/rejected": -0.05032349, "rewards/accuracies": 0.5, "rewards/margins": -0.00524902, "logps/rejected": -0.50292969, "logps/chosen": -0.55517578, "logits/rejected": -0.44628906, "logits/chosen": -1.12988281, "nll_loss": 0.55566406, "log_odds_ratio": -0.97314453, "log_odds_chosen": -0.33496094, "epoch": 0.39344262, "global_step/max_steps": "168/427", "percentage": "39.34%", "elapsed_time": "11h 41m 4s", "remaining_time": "18h 0m 50s"}
+{"loss": 0.43942261, "grad_norm": 0.0852342, "learning_rate": 3.057e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003999, "rewards/chosen": -0.03579712, "rewards/rejected": -0.50466919, "rewards/accuracies": 1.0, "rewards/margins": 0.46917725, "logps/rejected": -5.0390625, "logps/chosen": -0.35791016, "logits/rejected": -0.32128906, "logits/chosen": -0.61230469, "nll_loss": 0.35791016, "log_odds_ratio": -0.34716797, "log_odds_chosen": 5.30078125, "epoch": 0.39578454, "global_step/max_steps": "169/427", "percentage": "39.58%", "elapsed_time": "11h 43m 40s", "remaining_time": "17h 54m 14s"}
+{"loss": 0.43887329, "grad_norm": 0.09104977, "learning_rate": 3.045e-05, "memory(GiB)": 133.71, "train_speed(iter/s)": 0.003998, "rewards/chosen": -0.03686523, "rewards/rejected": -0.05151367, "rewards/accuracies": 0.5, "rewards/margins": 0.01464844, "logps/rejected": -0.51538086, "logps/chosen": -0.3684082, "logits/rejected": -0.26806641, "logits/chosen": -0.8203125, "nll_loss": 0.36889648, "log_odds_ratio": -0.66357422, "log_odds_chosen": 0.22485352, "epoch": 0.39812646, "global_step/max_steps": "170/427", "percentage": "39.81%", "elapsed_time": "11h 48m 3s", "remaining_time": "17h 50m 24s"}
+{"eval_loss": 0.46020046, "eval_runtime": 649.8011, "eval_samples_per_second": 0.649, "eval_steps_per_second": 0.082, "eval_rewards/chosen": -0.04084519, "eval_rewards/rejected": -0.10577969, "eval_rewards/accuracies": 0.6981132, "eval_rewards/margins": 0.06498286, "eval_logps/rejected": -1.05807781, "eval_logps/chosen": -0.40862691, "eval_logits/rejected": -0.30032292, "eval_logits/chosen": -0.79020488, "eval_nll_loss": 0.4086822, "eval_log_odds_ratio": -0.49878392, "eval_log_odds_chosen": 0.9986549, "epoch": 0.39812646, "global_step/max_steps": "170/427", "percentage": "39.81%", "elapsed_time": "11h 58m 52s", "remaining_time": "18h 6m 46s"}