Commit Graph
1698 Commits
Author SHA1 Message Date
affa284352 Refactor/reduce G/C/D/IoU if: else statements (#6087)
* Refactor the code to reduece else

* Update metrics.py

* Cleanup

Co-authored-by: Cmos <gen.chen@ubisoft.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-30 13:59:29 -08:00
Awsaf 7b6938d5b5 Log best results (#6085)
* log best result in summary

* comment added

* add space for `flake8`

* log `best/epoch`

* fix `dimension` for epoch

ValueError: all the input arrays must have same number of dimensions

* log `best/` in `utils.logger.__init__`

* fix pre-commit

1. missing whitespace around operator
2.  over-indented
2021-12-30 13:47:53 -08:00
Ayush ChaurasiaandGlenn Jocher 9155eb8641 W&B: Log best results after training ends (#6120)
* log best.pt metrics at train end

* update

* Update __init__.py

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-30 11:43:19 -08:00
Ayush ChaurasiaandGlenn Jocher db6ec66a60 W&B: track batch size after autobatch (#6039)
* track batch size after autobatch

* remove redundant import

* Update __init__.py

* Update __init__.py

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-23 14:23:50 +01:00
Deep Patel c72270c076 Init tensor directly on device (#6068)
Slightly more efficient than .to(device)
2021-12-23 13:49:00 +01:00
JieLiand李杰 afa5cfb0f8 Reduce G/D/CIoU logic operations (#6074)
Consider that the default value is CIOU,adjust the order of judgment could reduce the number of judgments.
And “elif CIoU:” didn't need 'if'.

Co-authored-by: 李杰 <360751194@qq.comqq.com>
2021-12-23 11:53:00 +01:00
Glenn Jocher 95c7bc25d3 OpenVINO Export (#6057)
* OpenVINO export

* Remove timeout

* Add 3 files

* str

* Constrain opset to 12

* Default ONNX opset to 12

* Make dir

* Make dir

* Cleanup

* Cleanup

* check_requirements(('openvino-dev',))
2021-12-22 20:29:48 +01:00
Glenn Jocher dc54ed5763 --freeze fix (#6044)
Fix for https://github.com/ultralytics/yolov5/issues/6038
2021-12-20 18:24:07 +01:00
Glenn Jocher b8a4babd60 Simplify set_logging() indexing (#6042) 2021-12-20 17:42:52 +01:00
Glenn Jocher 0db9d5b6a2 Kaggle LOGGER fix (#6041) 2021-12-20 17:30:46 +01:00
Glenn Jocher 26f0415287 Add dataset source citations (#6032) 2021-12-19 15:19:04 +01:00
Glenn Jocher abbdd4802e train -> val comment fix (#6024) 2021-12-17 16:43:00 +01:00
Felix YouandGlenn Jocher 361705d9be Multi-layer capable --freeze argument (#6019)
* support specfiy multiple frozen layers

* fix bug

* Cleanup Freeze section

* Cleanup argument

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-17 15:42:26 +01:00
Glenn Jocherandpre-commit-ci[bot] 407a905747 Check TensorRT>=8.0.0 version (#6021)
* Check TensorRT>=8.0.0 version

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-12-17 14:59:46 +01:00
Glenn Jocher c1249a47c7 *.torchscript inference self.jit fix (#6007) 2021-12-16 14:10:54 +01:00
Glenn Jocher 628817dfae Recommend jar xf file.zip for zips (#5993) 2021-12-15 17:19:19 +01:00
Glenn Jocher da9a1b719b Allow --weights URL (#5991) 2021-12-15 15:27:08 +01:00
Mrinal Jain b7d18f3636 Increase ar_thr from 20 to 100 for better detection on slender (high aspect ratio) objects (#5556)
* Making `ar_thr` available as a hyperparameter

* Disabling ar_thr as hyperparameter and computing from the dataset instead

* Fixing bug in ar_thr computation

* Fix `ar_thr` to 100
2021-12-15 14:42:23 +01:00
Glenn Jocher c9a46a60e0 Update callbacks.py with __init__() (#5979)
Add __init__() function.
2021-12-14 15:47:49 +01:00
Glenn Jocher d699c21c75 YOLOv5s6 params and FLOPs fix (#5977) 2021-12-14 11:24:39 +01:00
jinmcandGlenn Jocher 2d0c6afbfe make parameter ignore epochs (#5972)
* make parameter ignore epochs

ignore epochs functionality add to prevent spikes at the beginning when fitness spikes and decreases after.
Discussed at https://github.com/ultralytics/yolov5/issues/5971

* Update train.py

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-14 11:18:34 +01:00
Glenn Jocher e8ef8fb1ca pretrained=False fix (#5966)
* `pretriained=False` fix

Fix for https://github.com/ultralytics/yolov5/issues/5964

* CI speed improvement
2021-12-13 13:32:27 +01:00
Diego MontesandGlenn Jocher 19c56e60b1 Fix imgsz bug (#5948)
* fix imgsz bug

* Update detect.py

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-12 23:39:14 +01:00
8f354362cd Fix Detections class tolist() method (#5945)
* Fix tolist() to add the file for each Detection

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Fix PEP8 requirement for 2 spaces before an inline comment

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Cleanup

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-11 18:40:37 +01:00
Glenn Jocher 8f875d93a2 Refactor NUM_THREADS (#5954) 2021-12-11 14:32:18 +01:00
Diego MontesandGlenn Jocher 2c6317547a Add nms and agnostic nms to export.py (#5938)
* add nms and agnostic nms to export.py

* fix agnostic implies nms

* reorder args to group TF args

* PEP8 120 char

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-10 18:24:32 +01:00
iumyx2612 a42af30d8a Update strip_optimizer() (#5949)
Replace 'training_result' with 'best_fitness' in strip_optimizer() to match key with ckpt from train.py
2021-12-10 18:06:27 +01:00
Pascal MaillardandGlenn Jocher 922fbd8209 fix .gitignore not tracking existing folders (#5946)
* fix .gitignore not tracking existing folders

fix .gitignore so that the files that are in the repository are actually being tracked.

Everything in the data/ folder is ignored, which also means the subdirectories are ignored. Fix so that the subdirectories and their contents are still tracked.

* Remove data/trainings

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-10 17:32:09 +01:00
Felix YouandGlenn Jocher c45f9f678d Make select_device() robust to batch_size=-1 (#5940)
* Find out a bug. When set batch_size = -1 to use the autobatch.

reproduce:

* Fix type conflict

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-10 15:27:20 +01:00
Diego Montes 4fb6dd4b26 Fix ONNX opset inconsistency with parseargs and run args (#5937) 2021-12-09 23:10:16 +01:00
Glenn Jocher 5bdb28ed10 Default PyTorch Hub to autocast(False) (#5926) 2021-12-08 23:15:14 +01:00
Glenn Jocher c77a5a84e3 Absolute '/content/sample_data' (#5922) 2021-12-08 17:31:31 +01:00
Glenn Jocher a3d5f1d3e3 Revert "Update plot_lr_scheduler() (#5864)" (#5920)
This reverts commit 360eec6910.
2021-12-08 16:46:24 +01:00
Glenn Jocherandpre-commit-ci[bot] 7d56d45124 Add hardware checks to notebook_init() (#5919)
* Update notebook

* Update notebook

* update string

* update string

* Updates

* Updates

* Updates

* check both ipython and psutil

* remove sample_data if is_colab

* cleanup

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-12-08 14:57:03 +01:00
Glenn Jocher 581dc301a7 Add ONNX inference providers (#5918)
* Add ONNX inference providers

Fix for https://github.com/ultralytics/yolov5/issues/5916

* Update common.py
2021-12-08 13:37:33 +01:00
greg2451andGlenn Jocher 554f782537 Add *.engine (TensorRT extensions) to .gitignore (#5911)
* Add *.engine (TensorRT extensions) to .gitignore

* Update .dockerignore

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-07 18:20:16 +01:00
Ayush Chaurasia 3f152e5807 Update wandb_utils.py (#5908) 2021-12-07 16:09:11 +01:00
Li Zeng fa05f8c977 Detections().tolist() explicit argument fix (#5907)
debugged for missigned Detections attributes
2021-12-07 16:01:41 +01:00
Glenn Jocher 1075488d89 Single-command multiple-model export (#5882)
* Export multiple models in series

Export multiple models in series by adding additional `*.pt` files to the `--weights` argument, i.e.:

```bash
python export.py --include tflite --weights yolov5n.pt  # export 1 model
python export.py --include tflite --weights yolov5n.pt yolov5s.pt yolov5m.pt yolov5l.pt yolov5x.pt  # export 5 models
```

* Update export.py

* Update README.md
2021-12-04 16:28:40 +01:00
Glenn Jocher 7bf04d9bbf AutoShape() models as DetectMultiBackend() instances (#5845)
* Update AutoShape()

* autodownload ONNX

* Cleanup

* Finish updates

* Add Usage

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* fix device

* Update hubconf.py

* Update common.py

* smart param selection

* autodownload all formats

* autopad only pytorch models

* new_shape edits

* stride tensor fix

* Cleanup
2021-12-04 15:00:07 +01:00
Glenn Jocher d885799c71 Update nl after cutout() (#5873) 2021-12-03 15:28:14 +01:00
Can 360eec6910 Update plot_lr_scheduler() (#5864)
shallow copy modify originals
2021-12-03 13:37:45 +01:00
iumyx2612andGlenn Jocher 92a7391039 Add --workers 8 argument to val.py (#5857)
* Update val.py

Add an option to choose number of workers if not called by train.py

* Update comment

* 120 char line width

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-02 16:49:50 +01:00
Glenn Jocher 00e308f7be Update TorchScript suffix to *.torchscript (#5856) 2021-12-02 16:06:45 +01:00
Vishnu PradeepandGlenn Jocher 30db14fea8 Update val.py (#5838)
* Update val.py

Solving Non-ASCII character '\xf0' error during runtime

* Update val.py

Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2021-12-02 13:20:58 +01:00
gggmt 1679aacdc7 Update LoadImages ret_val=False handling (#5852)
Video errors may occur.
2021-12-02 10:57:39 +01:00
Yu Zhang e8f8f2b903 Avoid inplace modifyingimgs in LoadStreams (#5850)
When OpenCV retrieving image fail, original code would modify source images **inplace**, which may result in plotting bounding boxes on a black image. That is, before inference, source image `im0s[i]` is OK, but after inference before `Process predictions`,  `im0s[i]` may have been changed.
2021-12-02 10:51:19 +01:00
Glenn Jocher bc484579d7 Handle non-TTY wandb.errors.UsageError (#5839)
* `try: except (..., wandb.errors.UsageError)`

* bug fix
2021-12-01 15:38:02 +01:00
imyhxy a4207a202d Fix TensorRT potential unordered binding addresses (#5826)
* feat: change file suffix in pythonic way

* fix: enforce binding addresses order

* fix: enforce binding addresses order
2021-11-30 13:52:22 +01:00
Glenn Jocher 5ca5dd4c87 Update dataset_stats() to cv2.INTER_AREA (#5821) 2021-11-29 13:15:36 +01:00