nohup: ignoring input The following values were not passed to `accelerate launch` and had defaults used instead: More than one GPU was found, enabling multi-GPU training. If this was unintended please pass in `--num_processes=1`. `--num_machines` was set to a value of `1` `--mixed_precision` was set to a value of `'no'` `--dynamo_backend` was set to a value of `'no'` To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`. 07:42:46 INFO Loaded 34002 PRMTrainRecord rows from /workspace/fms4navigation/datasets/Faithfulness-Critic-Dataset/train_dataset.jsonl 07:42:46 INFO Loaded 512 PRMTrainRecord rows from /workspace/fms4navigation/datasets/Faithfulness-Critic-Dataset/val_dataset_512.jsonl 07:42:46 INFO Label balance (train, n=34002): 07:42:46 INFO overall CONSISTENT=14006 INCONSISTENT=19996 (41.2% pos) 07:42:46 INFO image_to_mj CONSISTENT=27995 INCONSISTENT= 6007 (82.3% pos) 07:42:46 INFO mj_to_action CONSISTENT=24708 INCONSISTENT= 9294 (72.7% pos) 07:42:46 INFO action_to_waypoints CONSISTENT=22682 INCONSISTENT=11320 (66.7% pos) 07:42:46 INFO mj_to_waypoints CONSISTENT=21033 INCONSISTENT=12969 (61.9% pos) 07:42:46 INFO Label balance (val, n=512): 07:42:46 INFO overall CONSISTENT= 209 INCONSISTENT= 303 (40.8% pos) 07:42:46 INFO image_to_mj CONSISTENT= 432 INCONSISTENT= 80 (84.4% pos) 07:42:46 INFO mj_to_action CONSISTENT= 364 INCONSISTENT= 148 (71.1% pos) 07:42:46 INFO action_to_waypoints CONSISTENT= 324 INCONSISTENT= 188 (63.3% pos) 07:42:46 INFO mj_to_waypoints CONSISTENT= 321 INCONSISTENT= 191 (62.7% pos) 07:42:46 INFO Loading processor: Qwen/Qwen3-VL-4B-Instruct Loading checkpoint shards: 0%| | 0/2 [00:00