MMIxaction2 动作检测测试失败，信号终止

问：

我正在使用 mmaction2 来测试一些动作识别和动作检测模型。 我在 kinetics400 上对许多可用的动作识别模型运行了 mani 测试，一切正常。

然后，我尝试在 AVA 数据集上对一些可用的动作检测模型（最慢，lfb）运行测试。不幸的是，在这些情况下，在所有测试周期结束时，执行都会失败。

(openmmlabGPU) daitan3@daitan3-workstation:~/decondDrive/pythonProjects/mmaction2$ python tools/test.py configs/detection/lfb/slowonly-lfb-nl_kinetics400-pretrained-r50_8xb12-4x16x1-20e_ava21-rgb.py https://download.openmmlab.com/mmaction/v1.0/detection/lfb/slowonly-lfb-nl_kinetics400-pretrained-r50_8xb12-4x16x1-20e_ava21-rgb/slowonly-lfb-nl_kinetics400-pretrained-r50_8xb12-4x16x1-20e_ava21-rgb_20220906-4c5b9f25.pth --dump result_lfb_nl.pkl >> result_lfb_nl.txt
Downloading: "https://download.openmmlab.com/mmaction/v1.0/detection/lfb/slowonly-lfb-nl_kinetics400-pretrained-r50_8xb12-4x16x1-20e_ava21-rgb/slowonly-lfb-nl_kinetics400-pretrained-r50_8xb12-4x16x1-20e_ava21-rgb_20220906-4c5b9f25.pth" to /home/daitan3/.cache/torch/hub/checkpoints/slowonly-lfb-nl_kinetics400-pretrained-r50_8xb12-4x16x1-20e_ava21-rgb_20220906-4c5b9f25.pth
100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 138M/138M [00:15<00:00, 9.03MB/s]

.
.
.
.


11/17 12:18:09 - mmengine - INFO - Epoch(test) [50140/50252]    eta: 0:00:03  time: 0.0450  data_time: 0.0016  memory: 1873  
11/17 12:18:10 - mmengine - INFO - Epoch(test) [50160/50252]    eta: 0:00:03  time: 0.0356  data_time: 0.0016  memory: 1873  
11/17 12:18:10 - mmengine - INFO - Epoch(test) [50180/50252]    eta: 0:00:02  time: 0.0317  data_time: 0.0016  memory: 1873  
11/17 12:18:11 - mmengine - INFO - Epoch(test) [50200/50252]    eta: 0:00:01  time: 0.0447  data_time: 0.0017  memory: 1873  
11/17 12:18:12 - mmengine - INFO - Epoch(test) [50220/50252]    eta: 0:00:01  time: 0.0438  data_time: 0.0016  memory: 1873  
11/17 12:18:13 - mmengine - INFO - Epoch(test) [50240/50252]    eta: 0:00:00  time: 0.0400  data_time: 0.0017  memory: 1873  
==> 1.19164 seconds to Reading GT results
==> 19.6223 seconds to Reading Detection results
Killed
(openmmlabGPU) daitan3@daitan3-workstation:~/decondDrive/pythonProjects/mmaction2$ Process ForkPoolWorker-9:
Process ForkPoolWorker-10:
Traceback (most recent call last):
Traceback (most recent call last):
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/pool.py", line 131, in worker
    put((job, i, result))
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/pool.py", line 131, in worker
    put((job, i, result))
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/queues.py", line 368, in put
    self._writer.send_bytes(obj)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 200, in send_bytes
    self._send_bytes(m[offset:offset + size])
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/queues.py", line 368, in put
    self._writer.send_bytes(obj)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 404, in _send_bytes
    self._send(header)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 368, in _send
    n = write(self._handle, buf)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 200, in send_bytes
    self._send_bytes(m[offset:offset + size])
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 404, in _send_bytes
    self._send(header)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 368, in _send
    n = write(self._handle, buf)
BrokenPipeError: [Errno 32] Broken pipe
BrokenPipeError: [Errno 32] Broken pipe

During handling of the above exception, another exception occurred:

Traceback (most recent call last):

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
    self.run()
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/pool.py", line 136, in worker
    put((job, i, (False, wrapped)))
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/queues.py", line 368, in put
    self._writer.send_bytes(obj)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 200, in send_bytes
    self._send_bytes(m[offset:offset + size])
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 404, in _send_bytes
    self._send(header)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 368, in _send
    n = write(self._handle, buf)
BrokenPipeError: [Errno 32] Broken pipe
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
    self.run()
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/pool.py", line 136, in worker
    put((job, i, (False, wrapped)))
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/queues.py", line 368, in put
    self._writer.send_bytes(obj)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 200, in send_bytes
    self._send_bytes(m[offset:offset + size])
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 404, in _send_bytes
    self._send(header)
  File "/home/daitan3/miniconda3/envs/openmmlabGPU/lib/python3.8/multiprocessing/connection.py", line 368, in _send
    n = write(self._handle, buf)
BrokenPipeError: [Errno 32] Broken pipe
(openmmlabGPU) daitan3@daitan3-workstation:~/decondDrive/pythonProjects/mmaction2$ 

我试图避免转储和输出重定向，但结果是一样的。

该测试已在 tux 下执行。在tux之外，它只会导致冻结需要重新启动的系统。

该系统是在具有 16 GB RAM 和 NVIDIA TITAN RTX 24 Gb 的 i7 上运行的 Ubuntu 20.04

有人可以帮我找到问题吗？