resume problem

[sezarxray]

Before submitting a bug report, please be aware that your issue must be reproducible with all of the following, otherwise it is non-actionable, and we can not help you:

• Current repo: 3e25f1e
• Common dataset: coco.yaml or coco128.yaml
  If this is a custom dataset/training question you must include your train*.jpg, test*.jpg and results.png figures, or we can not help you. You can generate these with utils.plot_results().

🐛 Bug

resume comment results in drop in accuracy

To Reproduce (REQUIRED)

python train.py --img 1024 --batch 16 --epochs 150 --data dataset_9910.yaml --weights yolov5m.pt --hyp hyp_evolved.yaml --workers 8 --name=p1 --resume

Environment

If applicable, add screenshots to help explain your problem.

• OS: [e.g. Ubuntu]
• GPU [e.g. 2080 Ti]

Additional context

Add any other context about the problem here.

[glenn-jocher]

@sezarxray --resume accepts zero additional arguments. The proper usage is:

python train.py --resume  # resumes from most recent training
python train.py --resume path/to/last.pt

[sezarxray]

thanks but even with you code problem exist!

[sezarxray]

before resume:

     Epoch   gpu_mem       box       obj       cls     total   targets  img_size
    17/149     7.45G   0.01701    0.0367 0.0006905   0.05439       116      1024:  89%|▉| 1108/1247 [1:45:07<13:11,  5.69^C

after resume:

     Epoch   gpu_mem       box       obj       cls     total   targets  img_size
    16/149     7.41G    0.1392   0.08824  0.009352    0.2368       114      1024:   1%|▏                             | 8/1247 [00:10<23:21,  1.13s/it]

[glenn-jocher]

@sezarxray --resume operates correctly. If you believe otherwise please file a full bug report with exact code to reproduce.

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.