mmdetection訓練時報錯：line 746, in print_map_summary label_names[j], num_gts[i, j], results[j][‘num_dets’],IndexError: list index out of range

使用mmdetection配置模型訓練自己的數據集時，會出現報錯

Traceback (most recent call last):File "tools/train.py", line 237, in <module>main()File "tools/train.py", line 233, in mainmeta=meta)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmdet-2.24.1-py3.7.egg/mmdet/apis/train.py", line 244, in train_detectorrunner.run(data_loaders, cfg.workflow)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 127, in runepoch_runner(data_loaders[i], **kwargs)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 54, in trainself.call_hook('after_train_epoch')File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmcv/runner/base_runner.py", line 309, in call_hookgetattr(hook, fn_name)(self)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmcv/runner/hooks/evaluation.py", line 267, in after_train_epochself._do_evaluate(runner)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmdet-2.24.1-py3.7.egg/mmdet/core/evaluation/eval_hooks.py", line 58, in _do_evaluatekey_score = self.evaluate(runner, results)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmcv/runner/hooks/evaluation.py", line 364, in evaluateresults, logger=runner.logger, **self.eval_kwargs)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmdet-2.24.1-py3.7.egg/mmdet/datasets/voc.py", line 92, in evaluateuse_legacy_coordinate=True)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmdet-2.24.1-py3.7.egg/mmdet/core/evaluation/mean_ap.py", line 683, in eval_mapmean_ap, eval_results, dataset, area_ranges, logger=logger)File "/home/liyinghao/anaconda3/envs/mmdet/lib/python3.7/site-packages/mmdet-2.24.1-py3.7.egg/mmdet/core/evaluation/mean_ap.py", line 746, in print_map_summarylabel_names[j], num_gts[i, j], results[j]['num_dets'], IndexError: list index out of range

此時是項目配置文件中num_classes沒有修改成自己的類別數

bbox_head=dict(type='YOLOXHead', num_classes=6, in_channels=128, feat_channels=128),train_cfg=dict(assigner=dict(type='SimOTAAssigner', center_radius=2.5)),

將num_classes改為自己的即可
每個人使用的配置文件不同，如果用pycham的話可以ctrl+f 搜索一下當前配置文件下的num_classes，進行修改

總結

以上是生活随笔為你收集整理的mmdetection踩坑记录的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。