* Autobatch

* fix mem

* fix mem2

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update train.py

* print result

* Cleanup print result

* swap fix in call

* to 64

* use total

* fix

* fix

* fix

* fix

* fix

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Cleanup printing

* Update final printout

* Update autobatch.py

* Update autobatch.py

* Update autobatch.py

* take EXIF orientation tags into account when fixing corrupt images

* fit 120 char

* sort imports

* Update local exif_transpose comment

We have a local inplace version that is faster than the official as the image is not copied. AutoShape() uses this for Hub models, but here it is not important as the datasets.py usage is infrequent (AutoShape() it is applied every image).

* Update datasets.py

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* take image files with uppercase extensions into account in autosplit

* case fix

* Refactor implementation

Removes additional variable (capital variable names are also only for global variables), and uses the same methodology as implemented earlier in datasets.py L409.

* Remove redundant rglob characters

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* Add `on_fit_epoch_end` callback

* Add results to train

* Update __init__.py

* Autofix duplicate labels

PR changes duplicate label handling from report error and ignore image-label pair to report warning and autofix image-label pair. 

This should fix this common issue for users and allow everyone to get started and get a model trained faster and easier than before.

* sign fix

* Cleanup

* Increment cache version

* all to any fix

* Improved check_suffix() robustness to `''` and `""`

* Cleanup

* fix dpp with wandb disabled

* PyCharm reformat

* Fix

* fix

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* fix dpp with wandb disabled

* PyCharm reformat

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* Add train class filter feature to datasets.py

Allows for training on a subset of total classes if `include_class` list is defined on datasets.py L448:
```python
        include_class = []  # filter labels to include only these classes (optional)
```

* segments fix

* refactor anchors and anchor_grid in Detect Layer

* fix CI failures by adding compatibility

* fix tf failure

* fix different devices errors

* Cleanup

* fix anchors overwriting issue

* better refactoring

* Remove self.anchor_grid shape check (redundant with self.grid check)

Also PEP8 / 120 line width

* Convert _make_grid() from static to dynamic method

* Remove anchor_grid.to(device)

clone() should already clone to same device as self.anchors

* fix different devices error

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* Re-open IP-camera videostream if disconnected. Provide IP-streaming analysis stability

* Re-open IP-camera videostream if disconnected. Provide IP-streaming analysis stability

* `self.imgs[i] *= 0` bug fix and cleanup

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* Fix E722, do not use bare 'except'

* Remove used codes

* Add FileNotFoundError in LoadImagesAndLabels

* Remove AssertionError

* Ignore LoadImagesAndLabels

* Ignore downloads.py

* Ignore torch_utils.py

* Ignore train.py

* Ignore datasets.py

* Enable utils/download.py

* Fixing exception in thop

* Remove unused code

* Fixing exception in LoadImagesAndLabels

* Fixing exception in exif_size

* Fixing exception in parse_model

* Ignore exceptions in requests

* Revert the exception as suggested

* Revert the exception as suggested

* added callbacks

* added back callback to main

* added save_dir to callback output

* merged in upstream

* removed ghost code

* fixed parsing error for google temp links

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

`is_valset_wandb_artifact` and `is_trainset_wandb_artifact` were referenced before assignment causing wandb to be unusable.

* fix isascii for python3.6

* update comment with python 3.7 note

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* Automatic Chinese fonts plotting

* Default PIL=False

`check_file()` is now limited to searching opt-in directories: /data, /models, /utils. This prevents large non-project directories like /.git and /venv from being searched, which may cause `check_file()` to slow significantly.

* Replace `os.system('unzip file.zip')` -> `ZipFile.extractall()`

* Cleanup

* Avoid out-of-image class labels

* Update plots.py

* Cleanup

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

* evolve fix

* Enable login timeout

* fix pkg

* check rank

* don't relogin

* fix

* reformat

* evolve fix

* Enable login timeout

* fix pkg

* check rank

* don't relogin

* Annotator `check_font()` RANK -1 remove progress

* Cleanup

* Simplify `check_requirements()` usage

* remove assert, print()

* Update `sys.path.append(str(ROOT))`

* Cleanup

* Refactor argparser printing to `print_args()`

* Cleanup

* evolve fix

* Enable login timeout

* fix pkg

PR should produce datasets sorted alphabetically by filename. Cache version incremented to 0.5. 

Note: will force a one-time re-caching of existing datasets on first-use.

* Unpin Pillow

* Update requirements.txt

* Update plots.py

* Fix val.py study plot

* call plot_val_study

* Rename plot_study_txt to plot_val_study

* added callbacks

* added back callback to main

* added save_dir to callback output

* merged in upstream

* removed ghost code

* added url check

* Add url2file()

* Update file-only

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>

Thank you for sharing nice open-source codes 👍

I applied to shuffle the order of all 4(or 9) images in mosaic augmentation

Currently, the order of images in mosaic augmentation is not completely random.
The remaining images except the first are randomly arranged. Apply shuffle all to increase the diversity of data composition.

* Initial commit

* Remove unused export_torchscript return

* ROOT variable

* Add prefix to fcn arg

* fix ROOT

* check_yaml into run()

* interim fixes

* imgsz=(320, 320)

* Hardcode tf_raw_resize False

* Finish opt elimination

* Update representative_dataset_gen()

* Update export.py with TF methods

* SiLU and GraphDef fixes

* file_size() directory handling feature

* export fixes

* add lambda: to representative_dataset

* Detect training False default

* Fuse false for TF models

* Embed agnostic NMS arguments

* Remove lambda

* TensorFlow.js export success

* Add pb to Usage

* Add *_tfjs_model/ to ignore files

* prepend YOLOv5 to function headers

* Remove end --- comments

* parameterize tfjs export pb file

* update run() data default /ROOT

* update --include help

* update imports

* return ct_model

* Consolidate TFLite export

* pb prerequisite to tfjs

* TF modules CamelCase

* Remove exports from tf.py and cleanup

* pass agnostic NMS arguments

* CI

* CI

* ignore *_web_model/

* Add tensorflow to CI dependencies

* CI tensorflow-cpu

* Update requirements.txt

* Remove tensorflow check_requirement

* CI coreml tfjs

* export only onnx torchscript

* reorder exports torchscript first

* Writeable test

* Fix

* Cleanup

* Fix `is_writeable()` for 3 OS support

* Update general.py