xiang-wuu/yolov5
Merge branch 'ultralytics:master' into main
fallback cpu, as no support for cuda device on hf spaces
due to build error changing cuda device tag
Add val.py no label warning (#8782)
requirement.txt updated to support gradio package
README updated to support integration of gradio & hugging spaces
inference and integration using gradio app
Fix confusion matrix update when no predictions are made (#8748)
Dynamic batch size support for TensorRT (#8526)
Merge original yolov5 master branch 'origin' into spaces main
initial commit
Revert protobuf<=3.20.1 (#8742)
fix: broken ``is_docker`` check (#8711)
Revert "Fix BGR->RGB Bug in albumentations #8641" (#8727)
coremltools>=5.2 for CoreML export (#8725)
Fix BGR->RGB Bug in albumentations #8641 (#8695)
New `HUBDatasetStats()` class (#8716)
Update dataloaders.py (#8714)
Avoid cv2 window init code on Windows (#8712)
Remove `else:` from load_image() (#8692)
Faster crop saving (#8696)
Move .dockerignore to root (#8690)
Set `torch.cuda.manual_seed_all()` for DDP (#8688)
Add generator and worker seed (#8602)
Update hubconf.py to reset logging level to INFO (#8680)
Update warning emojis (#8678)
Update hubconf.py to reset LOGGER.level after load (#8674)
Use contextlib's suppress method to silence an error (#8668)
Check TensorBoard logger before adding graph (#8664)
Upgrade onnxsim to v0.4.1 (#8632)
Fix `OMP_NUM_THREADS=1` for macOS (#8624)
assert torch!=1.12.0 for DDP training (#8621)
Refactor optimizer initialization (#8607)
FROM nvcr.io/nvidia/pytorch:22.06-py3 (#8600)
Link fuse() to AutoShape() for Hub models (#8599)
Add `pip install wheel` to avoid legacy `setup.py install` (#8597)
Update `tqdm>=4.64.0` and `thop>=0.1.1` (#8596)
Update `scipy>=1.7.3` (#8595)
Explicitly set `weight_decay` value (#8592)
Update requirements.txt comment spacing (#8562)
Assert `--optimize` not used with cuda device (#8569)
Raise error on suffix-less model path (#8561)
Add `thop>=0.1.0` (#8558)
Add TensorRT dependencies (#8553)
Update metrics.py with IoU protected divides (#8550)
CIoU protected divides (#8546)
val.py `assert ncm == nc` fix (#8545)
Fix `LoadImages()` with dataset YAML lists (#8517)
XML export `--half` fix (#8522)
Add `--half` arguments to export.py Usage examples (#8516)
