Skip to content

TRAIN_EVAL.md中的eval Bevformer指令有误 #265

Open

Description

@niouoklul

./tools/uniad_dist_train.sh ./projects/configs/bevformer/bevformer_base.py UniAD/ckpts/bevformer_r101_dcn_24ep.pth N_GPUS
应该改为
./tools/uniad_dist_eval.sh ./projects/configs/bevformer/base_bevformer.py ./ckpts/bevformer_r101_dcn_24ep.pth N_GPUS

Activity

  1. dd-xx-dot commented on Mar 8, 2026

    @dd-xx-dot

    I'm getting an error during evaluation:
    AssertionError: results must be a list from format_results in nuscenes_dataset.py.
    I've checked the return value of forward_test, and it's already wrapping the result in a list, like [bbox_results].
    However, the error still occurs.
    Have you successfully resolved this issue before? If so, what did you do?

    [>>>>>>>>>>>>>>>>>>>>>>>>>>] 6000/6000, 13.8 task/s, elapsed: 436s, ETA: 0s writing results to output/results.pkl Traceback (most recent call last): File "/root/workspace/UniAD/./tools/test.py", line 263, in <module> main() File "/root/workspace/UniAD/./tools/test.py", line 257, in main print(dataset.evaluate(outputs, **eval_kwargs)) File "/root/workspace/UniAD/venv/lib/python3.9/site-packages/mmdet3d/datasets/nuscenes_dataset.py", line 503, in evaluate result_files, tmp_dir = self.format_results(results, jsonfile_prefix) File "/root/workspace/UniAD/venv/lib/python3.9/site-packages/mmdet3d/datasets/nuscenes_dataset.py", line 443, in format_results assert isinstance(results, list), 'results must be a list' AssertionError: results must be a list ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 1) local_rank: 0 (pid: 285776) of binary: /root/workspace/UniAD/venv/bin/python Traceback (most recent call last): File "/root/workspace/UniAD/venv/bin/torchrun", line 7, in <module> sys.exit(main()) File "/root/workspace/UniAD/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper return f(*args, **kwargs) File "/root/workspace/UniAD/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 794, in main run(args) File "/root/workspace/UniAD/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 785, in run elastic_launch( File "/root/workspace/UniAD/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 134, in __call__ return launch_agent(self._config, self._entrypoint, list(args)) File "/root/workspace/UniAD/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 250, in launch_agent raise ChildFailedError( torch.distributed.elastic.multiprocessing.errors.ChildFailedError:

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions