Training in progress, step 30

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9c888db2569987e13d1c7cbbad0312c3a9366265d12698fc8268718d372c0d68
+oid sha256:0d64f2acf0ba695c086243b31f72620375545758a1b324a2b8856077ed488331
 size 6627156248

logging.jsonl

@@ -21,3 +21,14 @@
 {"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"}
+{"loss": 0.54748535, "grad_norm": 0.12414515, "learning_rate": 8.53e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004052, "rewards/chosen": -0.04528809, "rewards/rejected": -0.02444458, "rewards/accuracies": 0.0, "rewards/margins": -0.02084351, "logps/rejected": -0.24462891, "logps/chosen": -0.453125, "logits/rejected": -0.40209961, "logits/chosen": -0.796875, "nll_loss": 0.45166016, "log_odds_ratio": -1.12695312, "log_odds_chosen": -0.69238281, "epoch": 0.05152225, "global_step/max_steps": "22/1281", "percentage": "1.72%", "elapsed_time": "1h 30m 14s", "remaining_time": "3d 14h 4m 11s"}
+{"loss": 0.54998779, "grad_norm": 0.12814115, "learning_rate": 8.91e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004114, "rewards/chosen": -0.04998779, "rewards/rejected": -0.02371216, "rewards/accuracies": 0.0, "rewards/margins": -0.02627563, "logps/rejected": -0.23706055, "logps/chosen": -0.49951172, "logits/rejected": -0.37939453, "logits/chosen": -0.63964844, "nll_loss": 0.49804688, "log_odds_ratio": -1.18554688, "log_odds_chosen": -0.79443359, "epoch": 0.05386417, "global_step/max_steps": "23/1281", "percentage": "1.80%", "elapsed_time": "1h 32m 54s", "remaining_time": "3d 12h 41m 41s"}
+{"loss": 0.65063477, "grad_norm": 0.18243355, "learning_rate": 9.3e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004192, "rewards/chosen": -0.04476929, "rewards/rejected": -0.45700073, "rewards/accuracies": 0.25, "rewards/margins": 0.41278076, "logps/rejected": -4.57739258, "logps/chosen": -0.44726562, "logits/rejected": -0.49432373, "logits/chosen": -0.69580078, "nll_loss": 0.44628906, "log_odds_ratio": -0.8671875, "log_odds_chosen": 4.12890625, "epoch": 0.05620609, "global_step/max_steps": "24/1281", "percentage": "1.87%", "elapsed_time": "1h 35m 9s", "remaining_time": "3d 11h 3m 55s"}
+{"loss": 0.53256226, "grad_norm": 0.22813727, "learning_rate": 9.69e-06, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004223, "rewards/chosen": -0.03619385, "rewards/rejected": -0.32299805, "rewards/accuracies": 0.25, "rewards/margins": 0.28625488, "logps/rejected": -3.22216797, "logps/chosen": -0.36157227, "logits/rejected": -0.38720703, "logits/chosen": -0.69921875, "nll_loss": 0.36279297, "log_odds_ratio": -0.68164062, "log_odds_chosen": 3.08496094, "epoch": 0.05854801, "global_step/max_steps": "25/1281", "percentage": "1.95%", "elapsed_time": "1h 38m 24s", "remaining_time": "3d 10h 24m 19s"}
+{"loss": 0.5199585, "grad_norm": 0.15427108, "learning_rate": 1.008e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004306, "rewards/chosen": -0.04574585, "rewards/rejected": -0.02914429, "rewards/accuracies": 0.25, "rewards/margins": -0.01660156, "logps/rejected": -0.29150391, "logps/chosen": -0.45703125, "logits/rejected": -0.50683594, "logits/chosen": -0.72998047, "nll_loss": 0.45654297, "log_odds_ratio": -1.00488281, "log_odds_chosen": -0.51464844, "epoch": 0.06088993, "global_step/max_steps": "26/1281", "percentage": "2.03%", "elapsed_time": "1h 40m 21s", "remaining_time": "3d 8h 44m 34s"}
+{"loss": 0.51290894, "grad_norm": 0.14327918, "learning_rate": 1.047e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004361, "rewards/chosen": -0.05859375, "rewards/rejected": -0.02770996, "rewards/accuracies": 0.0, "rewards/margins": -0.03091431, "logps/rejected": -0.27709961, "logps/chosen": -0.58642578, "logits/rejected": -0.41548157, "logits/chosen": -0.6640625, "nll_loss": 0.58691406, "log_odds_ratio": -1.25585938, "log_odds_chosen": -0.90136719, "epoch": 0.06323185, "global_step/max_steps": "27/1281", "percentage": "2.11%", "elapsed_time": "1h 42m 55s", "remaining_time": "3d 7h 40m 13s"}
+{"loss": 0.58557129, "grad_norm": 0.12187042, "learning_rate": 1.085e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004384, "rewards/chosen": -0.04632568, "rewards/rejected": -0.02890015, "rewards/accuracies": 0.0, "rewards/margins": -0.01742554, "logps/rejected": -0.28955078, "logps/chosen": -0.46289062, "logits/rejected": -0.28991699, "logits/chosen": -0.48071289, "nll_loss": 0.46240234, "log_odds_ratio": -1.04101562, "log_odds_chosen": -0.55957031, "epoch": 0.06557377, "global_step/max_steps": "28/1281", "percentage": "2.19%", "elapsed_time": "1h 46m 11s", "remaining_time": "3d 7h 12m 10s"}
+{"loss": 0.53582764, "grad_norm": 0.11856092, "learning_rate": 1.124e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004414, "rewards/chosen": -0.02923584, "rewards/rejected": -0.19833374, "rewards/accuracies": 0.25, "rewards/margins": 0.16958618, "logps/rejected": -1.97949219, "logps/chosen": -0.29248047, "logits/rejected": -0.25146484, "logits/chosen": -0.41552734, "nll_loss": 0.29248047, "log_odds_ratio": -0.6640625, "log_odds_chosen": 1.86157227, "epoch": 0.06791569, "global_step/max_steps": "29/1281", "percentage": "2.26%", "elapsed_time": "1h 49m 14s", "remaining_time": "3d 6h 36m 10s"}
+{"loss": 0.5987854, "grad_norm": 0.12445131, "learning_rate": 1.163e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004477, "rewards/chosen": -0.03991699, "rewards/rejected": -0.40048218, "rewards/accuracies": 0.25, "rewards/margins": 0.36141968, "logps/rejected": -4.00097656, "logps/chosen": -0.39941406, "logits/rejected": -0.20629883, "logits/chosen": -0.38500977, "nll_loss": 0.39941406, "log_odds_ratio": -0.67480469, "log_odds_chosen": 3.7109375, "epoch": 0.07025761, "global_step/max_steps": "30/1281", "percentage": "2.34%", "elapsed_time": "1h 51m 24s", "remaining_time": "3d 5h 25m 55s"}
+{"eval_loss": 0.56994921, "eval_runtime": 642.2195, "eval_samples_per_second": 0.657, "eval_steps_per_second": 0.083, "eval_rewards/chosen": -0.04699477, "eval_rewards/rejected": -0.08173529, "eval_rewards/accuracies": 0.13207547, "eval_rewards/margins": 0.03477852, "eval_logps/rejected": -0.81732756, "eval_logps/chosen": -0.47000295, "eval_logits/rejected": -0.2889151, "eval_logits/chosen": -0.53823537, "eval_nll_loss": 0.46996608, "eval_log_odds_ratio": -1.02137387, "eval_log_odds_chosen": 0.02808078, "epoch": 0.07025761, "global_step/max_steps": "30/1281", "percentage": "2.34%", "elapsed_time": "2h 2m 7s", "remaining_time": "3d 12h 52m 15s"}
+{"loss": 0.47167969, "grad_norm": 0.12563924, "learning_rate": 1.202e-05, "memory(GiB)": 133.6, "train_speed(iter/s)": 0.004117, "rewards/chosen": -0.04330444, "rewards/rejected": -0.02688599, "rewards/accuracies": 0.25, "rewards/margins": -0.01641846, "logps/rejected": -0.26879883, "logps/chosen": -0.43261719, "logits/rejected": -0.22250366, "logits/chosen": -0.46716309, "nll_loss": 0.43310547, "log_odds_ratio": -1.02441406, "log_odds_chosen": -0.53173828, "epoch": 0.07259953, "global_step/max_steps": "31/1281", "percentage": "2.42%", "elapsed_time": "2h 5m 14s", "remaining_time": "3d 12h 10m 23s"}