Repository: EASY-EAI/yolov5 Branch: master Commit: 4e14e074579f Files: 102 Total size: 588.6 KB Directory structure: gitextract_19tvehv_/ ├── .dockerignore ├── .gitattributes ├── .github/ │ ├── ISSUE_TEMPLATE/ │ │ ├── bug-report.md │ │ ├── feature-request.md │ │ └── question.md │ ├── dependabot.yml │ └── workflows/ │ ├── ci-testing.yml │ ├── codeql-analysis.yml │ ├── greetings.yml │ ├── rebase.yml │ └── stale.yml ├── .gitignore ├── .pre-commit-config.yaml ├── CONTRIBUTING.md ├── Dockerfile ├── LICENSE ├── README.md ├── data/ │ ├── Argoverse.yaml │ ├── GlobalWheat2020.yaml │ ├── Objects365.yaml │ ├── SKU-110K.yaml │ ├── VisDrone.yaml │ ├── coco.yaml │ ├── coco128.yaml │ ├── hyps/ │ │ ├── hyp.finetune.yaml │ │ ├── hyp.finetune_objects365.yaml │ │ ├── hyp.scratch-high.yaml │ │ ├── hyp.scratch-low.yaml │ │ ├── hyp.scratch-med.yaml │ │ └── hyp.scratch.yaml │ ├── mask.yaml │ ├── scripts/ │ │ ├── download_weights.sh │ │ ├── get_coco.sh │ │ └── get_coco128.sh │ ├── voc.yaml │ └── xView.yaml ├── detect.py ├── export.py ├── hubconf.py ├── models/ │ ├── __init__.py │ ├── common.py │ ├── common_rk_plug_in.py │ ├── experimental.py │ ├── hub/ │ │ ├── anchors.yaml │ │ ├── yolov3-spp.yaml │ │ ├── yolov3-tiny.yaml │ │ ├── yolov3.yaml │ │ ├── yolov5-bifpn.yaml │ │ ├── yolov5-fpn.yaml │ │ ├── yolov5-p2.yaml │ │ ├── yolov5-p6.yaml │ │ ├── yolov5-p7.yaml │ │ ├── yolov5-panet.yaml │ │ ├── yolov5l6.yaml │ │ ├── yolov5m6.yaml │ │ ├── yolov5n6.yaml │ │ ├── yolov5s-ghost.yaml │ │ ├── yolov5s-transformer.yaml │ │ ├── yolov5s6.yaml │ │ └── yolov5x6.yaml │ ├── tf.py │ ├── yolo.py │ ├── yolov5l.yaml │ ├── yolov5m.yaml │ ├── yolov5n.yaml │ ├── yolov5s.yaml │ └── yolov5x.yaml ├── requirements.txt ├── setup.cfg ├── train.py ├── tutorial.ipynb ├── utils/ │ ├── __init__.py │ ├── activations.py │ ├── augmentations.py │ ├── autoanchor.py │ ├── autobatch.py │ ├── aws/ │ │ ├── __init__.py │ │ ├── mime.sh │ │ ├── resume.py │ │ └── userdata.sh │ ├── callbacks.py │ ├── datasets.py │ ├── downloads.py │ ├── flask_rest_api/ │ │ ├── README.md │ │ ├── example_request.py │ │ └── restapi.py │ ├── general.py │ ├── google_app_engine/ │ │ ├── Dockerfile │ │ ├── additional_requirements.txt │ │ └── app.yaml │ ├── loggers/ │ │ ├── __init__.py │ │ └── wandb/ │ │ ├── README.md │ │ ├── __init__.py │ │ ├── log_dataset.py │ │ ├── sweep.py │ │ ├── sweep.yaml │ │ └── wandb_utils.py │ ├── loss.py │ ├── metrics.py │ ├── plots.py │ └── torch_utils.py └── val.py ================================================ FILE CONTENTS ================================================ ================================================ FILE: .dockerignore ================================================ # Repo-specific DockerIgnore ------------------------------------------------------------------------------------------- #.git .cache .idea runs output coco storage.googleapis.com data/samples/* **/results*.csv *.jpg # Neural Network weights ----------------------------------------------------------------------------------------------- **/*.pt **/*.pth **/*.onnx **/*.engine **/*.mlmodel **/*.torchscript **/*.torchscript.pt **/*.tflite **/*.h5 **/*.pb *_saved_model/ *_web_model/ # Below Copied From .gitignore ----------------------------------------------------------------------------------------- # Below Copied From .gitignore ----------------------------------------------------------------------------------------- # GitHub Python GitIgnore ---------------------------------------------------------------------------------------------- # Byte-compiled / optimized / DLL files __pycache__/ *.py[cod] *$py.class # C extensions *.so # Distribution / packaging .Python env/ build/ develop-eggs/ dist/ downloads/ eggs/ .eggs/ lib/ lib64/ parts/ sdist/ var/ wheels/ *.egg-info/ wandb/ .installed.cfg *.egg # PyInstaller # Usually these files are written by a python script from a template # before PyInstaller builds the exe, so as to inject date/other infos into it. *.manifest *.spec # Installer logs pip-log.txt pip-delete-this-directory.txt # Unit test / coverage reports htmlcov/ .tox/ .coverage .coverage.* .cache nosetests.xml coverage.xml *.cover .hypothesis/ # Translations *.mo *.pot # Django stuff: *.log local_settings.py # Flask stuff: instance/ .webassets-cache # Scrapy stuff: .scrapy # Sphinx documentation docs/_build/ # PyBuilder target/ # Jupyter Notebook .ipynb_checkpoints # pyenv .python-version # celery beat schedule file celerybeat-schedule # SageMath parsed files *.sage.py # dotenv .env # virtualenv .venv* venv*/ ENV*/ # Spyder project settings .spyderproject .spyproject # Rope project settings .ropeproject # mkdocs documentation /site # mypy .mypy_cache/ # https://github.com/github/gitignore/blob/master/Global/macOS.gitignore ----------------------------------------------- # General .DS_Store .AppleDouble .LSOverride # Icon must end with two \r Icon Icon? # Thumbnails ._* # Files that might appear in the root of a volume .DocumentRevisions-V100 .fseventsd .Spotlight-V100 .TemporaryItems .Trashes .VolumeIcon.icns .com.apple.timemachine.donotpresent # Directories potentially created on remote AFP share .AppleDB .AppleDesktop Network Trash Folder Temporary Items .apdisk # https://github.com/github/gitignore/blob/master/Global/JetBrains.gitignore # Covers JetBrains IDEs: IntelliJ, RubyMine, PhpStorm, AppCode, PyCharm, CLion, Android Studio and WebStorm # Reference: https://intellij-support.jetbrains.com/hc/en-us/articles/206544839 # User-specific stuff: .idea/* .idea/**/workspace.xml .idea/**/tasks.xml .idea/dictionaries .html # Bokeh Plots .pg # TensorFlow Frozen Graphs .avi # videos # Sensitive or high-churn files: .idea/**/dataSources/ .idea/**/dataSources.ids .idea/**/dataSources.local.xml .idea/**/sqlDataSources.xml .idea/**/dynamic.xml .idea/**/uiDesigner.xml # Gradle: .idea/**/gradle.xml .idea/**/libraries # CMake cmake-build-debug/ cmake-build-release/ # Mongo Explorer plugin: .idea/**/mongoSettings.xml ## File-based project format: *.iws ## Plugin-specific files: # IntelliJ out/ # mpeltonen/sbt-idea plugin .idea_modules/ # JIRA plugin atlassian-ide-plugin.xml # Cursive Clojure plugin .idea/replstate.xml # Crashlytics plugin (for Android Studio and IntelliJ) com_crashlytics_export_strings.xml crashlytics.properties crashlytics-build.properties fabric.properties ================================================ FILE: .gitattributes ================================================ # this drop notebooks from GitHub language stats *.ipynb linguist-vendored ================================================ FILE: .github/ISSUE_TEMPLATE/bug-report.md ================================================ --- name: "🐛 Bug report" about: Create a report to help us improve title: '' labels: bug assignees: '' --- 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**: run `git fetch && git status -uno` to check and `git pull` to update repo - **Common dataset**: coco.yaml or coco128.yaml - **Common environment**: Colab, Google Cloud, or Docker image. See https://github.com/ultralytics/yolov5#environments 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 A clear and concise description of what the bug is. ## To Reproduce (REQUIRED) Input: ``` import torch a = torch.tensor([5]) c = a / 0 ``` Output: ``` Traceback (most recent call last): File "/Users/glennjocher/opt/anaconda3/envs/env1/lib/python3.7/site-packages/IPython/core/interactiveshell.py", line 3331, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "", line 5, in c = a / 0 RuntimeError: ZeroDivisionError ``` ## Expected behavior A clear and concise description of what you expected to happen. ## 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. ================================================ FILE: .github/ISSUE_TEMPLATE/feature-request.md ================================================ --- name: "🚀 Feature request" about: Suggest an idea for this project title: '' labels: enhancement assignees: '' --- ## 🚀 Feature ## Motivation ## Pitch ## Alternatives ## Additional context ================================================ FILE: .github/ISSUE_TEMPLATE/question.md ================================================ --- name: "❓Question" about: Ask a general question title: '' labels: question assignees: '' --- ## ❔Question ## Additional context ================================================ FILE: .github/dependabot.yml ================================================ version: 2 updates: - package-ecosystem: pip directory: "/" schedule: interval: weekly time: "04:00" open-pull-requests-limit: 10 reviewers: - glenn-jocher labels: - dependencies ================================================ FILE: .github/workflows/ci-testing.yml ================================================ name: CI CPU testing on: # https://help.github.com/en/actions/reference/events-that-trigger-workflows push: branches: [ master ] pull_request: # The branches below must be a subset of the branches above branches: [ master ] schedule: - cron: '0 0 * * *' # Runs at 00:00 UTC every day jobs: cpu-tests: runs-on: ${{ matrix.os }} strategy: fail-fast: false matrix: os: [ubuntu-latest, macos-latest, windows-latest] python-version: [3.8] model: ['yolov5s'] # models to test # Timeout: https://stackoverflow.com/a/59076067/4521646 timeout-minutes: 50 steps: - uses: actions/checkout@v2 - name: Set up Python ${{ matrix.python-version }} uses: actions/setup-python@v2 with: python-version: ${{ matrix.python-version }} # Note: This uses an internal pip API and may not always work # https://github.com/actions/cache/blob/master/examples.md#multiple-oss-in-a-workflow - name: Get pip cache id: pip-cache run: | python -c "from pip._internal.locations import USER_CACHE_DIR; print('::set-output name=dir::' + USER_CACHE_DIR)" - name: Cache pip uses: actions/cache@v1 with: path: ${{ steps.pip-cache.outputs.dir }} key: ${{ runner.os }}-${{ matrix.python-version }}-pip-${{ hashFiles('requirements.txt') }} restore-keys: | ${{ runner.os }}-${{ matrix.python-version }}-pip- - name: Install dependencies run: | python -m pip install --upgrade pip pip install -qr requirements.txt -f https://download.pytorch.org/whl/cpu/torch_stable.html pip install -q onnx python --version pip --version pip list shell: bash - name: Download data run: | # curl -L -o tmp.zip https://github.com/ultralytics/yolov5/releases/download/v1.0/coco128.zip # unzip -q tmp.zip -d ../ # rm tmp.zip - name: Tests workflow run: | # export PYTHONPATH="$PWD" # to run '$ python *.py' files in subdirectories di=cpu # inference devices # define device # train python train.py --img 128 --batch 16 --weights weights/${{ matrix.model }}.pt --cfg models/${{ matrix.model }}.yaml --epochs 1 --device $di # detect python detect.py --weights weights/${{ matrix.model }}.pt --device $di python detect.py --weights runs/train/exp/weights/last.pt --device $di # test python test.py --img 128 --batch 16 --weights weights/${{ matrix.model }}.pt --device $di python test.py --img 128 --batch 16 --weights runs/train/exp/weights/last.pt --device $di python hubconf.py # hub python models/yolo.py --cfg models/${{ matrix.model }}.yaml # inspect python models/export.py --img 128 --batch 1 --weights weights/${{ matrix.model }}.pt # export shell: bash ================================================ FILE: .github/workflows/codeql-analysis.yml ================================================ # This action runs GitHub's industry-leading static analysis engine, CodeQL, against a repository's source code to find security vulnerabilities. # https://github.com/github/codeql-action name: "CodeQL" on: schedule: - cron: '0 0 1 * *' # Runs at 00:00 UTC on the 1st of every month jobs: analyze: name: Analyze runs-on: ubuntu-latest strategy: fail-fast: false matrix: language: [ 'python' ] # CodeQL supports [ 'cpp', 'csharp', 'go', 'java', 'javascript', 'python' ] # Learn more: # https://docs.github.com/en/free-pro-team@latest/github/finding-security-vulnerabilities-and-errors-in-your-code/configuring-code-scanning#changing-the-languages-that-are-analyzed steps: - name: Checkout repository uses: actions/checkout@v2 # Initializes the CodeQL tools for scanning. - name: Initialize CodeQL uses: github/codeql-action/init@v1 with: languages: ${{ matrix.language }} # If you wish to specify custom queries, you can do so here or in a config file. # By default, queries listed here will override any specified in a config file. # Prefix the list here with "+" to use these queries and those in the config file. # queries: ./path/to/local/query, your-org/your-repo/queries@main # Autobuild attempts to build any compiled languages (C/C++, C#, or Java). # If this step fails, then you should remove it and run the build manually (see below) - name: Autobuild uses: github/codeql-action/autobuild@v1 # ℹ️ Command-line programs to run using the OS shell. # 📚 https://git.io/JvXDl # ✏️ If the Autobuild fails above, remove it and uncomment the following three lines # and modify them (or add more) to build your code if your project # uses a compiled language #- run: | # make bootstrap # make release - name: Perform CodeQL Analysis uses: github/codeql-action/analyze@v1 ================================================ FILE: .github/workflows/greetings.yml ================================================ name: Greetings on: [pull_request_target, issues] jobs: greeting: runs-on: ubuntu-latest steps: - uses: actions/first-interaction@v1 with: repo-token: ${{ secrets.GITHUB_TOKEN }} pr-message: | 👋 Hello @${{ github.actor }}, thank you for submitting a 🚀 PR! To allow your work to be integrated as seamlessly as possible, we advise you to: - ✅ Verify your PR is **up-to-date with origin/master.** If your PR is behind origin/master an automatic [GitHub actions](https://github.com/ultralytics/yolov5/blob/master/.github/workflows/rebase.yml) rebase may be attempted by including the /rebase command in a comment body, or by running the following code, replacing 'feature' with the name of your local branch: ```bash git remote add upstream https://github.com/ultralytics/yolov5.git git fetch upstream git checkout feature # <----- replace 'feature' with local branch name git rebase upstream/master git push -u origin -f ``` - ✅ Verify all Continuous Integration (CI) **checks are passing**. - ✅ Reduce changes to the absolute **minimum** required for your bug fix or feature addition. _"It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is."_ -Bruce Lee issue-message: | 👋 Hello @${{ github.actor }}, thank you for your interest in 🚀 YOLOv5! Please visit our ⭐️ [Tutorials](https://github.com/ultralytics/yolov5/wiki#tutorials) to get started, where you can find quickstart guides for simple tasks like [Custom Data Training](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data) all the way to advanced concepts like [Hyperparameter Evolution](https://github.com/ultralytics/yolov5/issues/607). If this is a 🐛 Bug Report, please provide screenshots and **minimum viable code to reproduce your issue**, otherwise we can not help you. If this is a custom training ❓ Question, please provide as much information as possible, including dataset images, training logs, screenshots, and a public link to online [W&B logging](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data#visualize) if available. For business inquiries or professional support requests please visit https://www.ultralytics.com or email Glenn Jocher at glenn.jocher@ultralytics.com. ## Requirements Python 3.8 or later with all [requirements.txt](https://github.com/ultralytics/yolov5/blob/master/requirements.txt) dependencies installed, including `torch>=1.7`. To install run: ```bash $ pip install -r requirements.txt ``` ## Environments YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including [CUDA](https://developer.nvidia.com/cuda)/[CUDNN](https://developer.nvidia.com/cudnn), [Python](https://www.python.org/) and [PyTorch](https://pytorch.org/) preinstalled): - **Google Colab and Kaggle** notebooks with free GPU: Open In Colab Open In Kaggle - **Google Cloud** Deep Learning VM. See [GCP Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/GCP-Quickstart) - **Amazon** Deep Learning AMI. See [AWS Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/AWS-Quickstart) - **Docker Image**. See [Docker Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/Docker-Quickstart) Docker Pulls ## Status ![CI CPU testing](https://github.com/ultralytics/yolov5/workflows/CI%20CPU%20testing/badge.svg) If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), testing ([test.py](https://github.com/ultralytics/yolov5/blob/master/test.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/models/export.py)) on MacOS, Windows, and Ubuntu every 24 hours and on every commit. ================================================ FILE: .github/workflows/rebase.yml ================================================ name: Automatic Rebase # https://github.com/marketplace/actions/automatic-rebase on: issue_comment: types: [created] jobs: rebase: name: Rebase if: github.event.issue.pull_request != '' && contains(github.event.comment.body, '/rebase') runs-on: ubuntu-latest steps: - name: Checkout the latest code uses: actions/checkout@v2 with: fetch-depth: 0 - name: Automatic Rebase uses: cirrus-actions/rebase@1.3.1 env: GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }} ================================================ FILE: .github/workflows/stale.yml ================================================ name: Close stale issues on: schedule: - cron: "0 0 * * *" jobs: stale: runs-on: ubuntu-latest steps: - uses: actions/stale@v3 with: repo-token: ${{ secrets.GITHUB_TOKEN }} stale-issue-message: 'This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.' stale-pr-message: 'This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.' days-before-stale: 30 days-before-close: 5 exempt-issue-labels: 'documentation,tutorial' operations-per-run: 100 # The maximum number of operations per run, used to control rate limiting. ================================================ FILE: .gitignore ================================================ # Repo-specific GitIgnore ---------------------------------------------------------------------------------------------- *.tif *.tiff *.heic *.TIF *.TIFF *.HEIC *.mp4 *.mov *.MOV *.avi *.data *.json *.cfg !setup.cfg !cfg/yolov3*.cfg storage.googleapis.com runs/* data/* !data/images !data/*.yaml !data/hyps !data/scripts !data/images !data/images/zidane.jpg !data/images/bus.jpg !data/*.sh results*.csv # Datasets ------------------------------------------------------------------------------------------------------------- coco/ coco128/ VOC/ # MATLAB GitIgnore ----------------------------------------------------------------------------------------------------- *.m~ *.mat !targets*.mat # Neural Network weights ----------------------------------------------------------------------------------------------- *.weights *.pt *.pb *.onnx *.engine *.mlmodel *.torchscript *.tflite *.h5 *_saved_model/ *_web_model/ darknet53.conv.74 yolov3-tiny.conv.15 # GitHub Python GitIgnore ---------------------------------------------------------------------------------------------- # Byte-compiled / optimized / DLL files __pycache__/ *.py[cod] *$py.class # C extensions *.so # Distribution / packaging .Python env/ build/ develop-eggs/ dist/ downloads/ eggs/ .eggs/ lib/ lib64/ parts/ sdist/ var/ wheels/ *.egg-info/ /wandb/ .installed.cfg *.egg # PyInstaller # Usually these files are written by a python script from a template # before PyInstaller builds the exe, so as to inject date/other infos into it. *.manifest *.spec # Installer logs pip-log.txt pip-delete-this-directory.txt # Unit test / coverage reports htmlcov/ .tox/ .coverage .coverage.* .cache nosetests.xml coverage.xml *.cover .hypothesis/ # Translations *.mo *.pot # Django stuff: *.log local_settings.py # Flask stuff: instance/ .webassets-cache # Scrapy stuff: .scrapy # Sphinx documentation docs/_build/ # PyBuilder target/ # Jupyter Notebook .ipynb_checkpoints # pyenv .python-version # celery beat schedule file celerybeat-schedule # SageMath parsed files *.sage.py # dotenv .env # virtualenv .venv* venv*/ ENV*/ # Spyder project settings .spyderproject .spyproject # Rope project settings .ropeproject # mkdocs documentation /site # mypy .mypy_cache/ # https://github.com/github/gitignore/blob/master/Global/macOS.gitignore ----------------------------------------------- # General .DS_Store .AppleDouble .LSOverride # Icon must end with two \r Icon Icon? # Thumbnails ._* # Files that might appear in the root of a volume .DocumentRevisions-V100 .fseventsd .Spotlight-V100 .TemporaryItems .Trashes .VolumeIcon.icns .com.apple.timemachine.donotpresent # Directories potentially created on remote AFP share .AppleDB .AppleDesktop Network Trash Folder Temporary Items .apdisk # https://github.com/github/gitignore/blob/master/Global/JetBrains.gitignore # Covers JetBrains IDEs: IntelliJ, RubyMine, PhpStorm, AppCode, PyCharm, CLion, Android Studio and WebStorm # Reference: https://intellij-support.jetbrains.com/hc/en-us/articles/206544839 # User-specific stuff: .idea/* .idea/**/workspace.xml .idea/**/tasks.xml .idea/dictionaries .html # Bokeh Plots .pg # TensorFlow Frozen Graphs .avi # videos # Sensitive or high-churn files: .idea/**/dataSources/ .idea/**/dataSources.ids .idea/**/dataSources.local.xml .idea/**/sqlDataSources.xml .idea/**/dynamic.xml .idea/**/uiDesigner.xml # Gradle: .idea/**/gradle.xml .idea/**/libraries # CMake cmake-build-debug/ cmake-build-release/ # Mongo Explorer plugin: .idea/**/mongoSettings.xml ## File-based project format: *.iws ## Plugin-specific files: # IntelliJ out/ # mpeltonen/sbt-idea plugin .idea_modules/ # JIRA plugin atlassian-ide-plugin.xml # Cursive Clojure plugin .idea/replstate.xml # Crashlytics plugin (for Android Studio and IntelliJ) com_crashlytics_export_strings.xml crashlytics.properties crashlytics-build.properties fabric.properties ================================================ FILE: .pre-commit-config.yaml ================================================ # Define hooks for code formations # Will be applied on any updated commit files if a user has installed and linked commit hook default_language_version: python: python3.8 # Define bot property if installed via https://github.com/marketplace/pre-commit-ci ci: autofix_prs: true autoupdate_commit_msg: '[pre-commit.ci] pre-commit suggestions' autoupdate_schedule: quarterly # submodules: true repos: - repo: https://github.com/pre-commit/pre-commit-hooks rev: v4.0.1 hooks: - id: end-of-file-fixer - id: trailing-whitespace - id: check-case-conflict - id: check-yaml - id: check-toml - id: pretty-format-json - id: check-docstring-first - repo: https://github.com/asottile/pyupgrade rev: v2.23.1 hooks: - id: pyupgrade args: [--py36-plus] name: Upgrade code - repo: https://github.com/PyCQA/isort rev: 5.9.3 hooks: - id: isort name: Sort imports # TODO #- repo: https://github.com/pre-commit/mirrors-yapf # rev: v0.31.0 # hooks: # - id: yapf # name: formatting # TODO #- repo: https://github.com/executablebooks/mdformat # rev: 0.7.7 # hooks: # - id: mdformat # additional_dependencies: # - mdformat-gfm # - mdformat-black # - mdformat_frontmatter # TODO #- repo: https://github.com/asottile/yesqa # rev: v1.2.3 # hooks: # - id: yesqa - repo: https://github.com/PyCQA/flake8 rev: 3.9.2 hooks: - id: flake8 name: PEP8 ================================================ FILE: CONTRIBUTING.md ================================================ ## Contributing to YOLOv5 🚀 We love your input! We want to make contributing to YOLOv5 as easy and transparent as possible, whether it's: - Reporting a bug - Discussing the current state of the code - Submitting a fix - Proposing a new feature - Becoming a maintainer YOLOv5 works so well due to our combined community effort, and for every small improvement you contribute you will be helping push the frontiers of what's possible in AI 😃! ## Submitting a Pull Request (PR) 🛠️ Submitting a PR is easy! This example shows how to submit a PR for updating `requirements.txt` in 4 steps: ### 1. Select File to Update Select `requirements.txt` to update by clicking on it in GitHub.

PR_step1

### 2. Click 'Edit this file' Button is in top-right corner.

PR_step2

### 3. Make Changes Change `matplotlib` version from `3.2.2` to `3.3`.

PR_step3

### 4. Preview Changes and Submit PR Click on the **Preview changes** tab to verify your updates. At the bottom of the screen select 'Create a **new branch** for this commit', assign your branch a descriptive name such as `fix/matplotlib_version` and click the green **Propose changes** button. All done, your PR is now submitted to YOLOv5 for review and approval 😃!

PR_step4

### PR recommendations To allow your work to be integrated as seamlessly as possible, we advise you to: - ✅ Verify your PR is **up-to-date with upstream/master.** If your PR is behind upstream/master an automatic [GitHub actions](https://github.com/ultralytics/yolov5/blob/master/.github/workflows/rebase.yml) rebase may be attempted by including the /rebase command in a comment body, or by running the following code, replacing 'feature' with the name of your local branch: ```bash git remote add upstream https://github.com/ultralytics/yolov5.git git fetch upstream git checkout feature # <----- replace 'feature' with local branch name git merge upstream/master git push -u origin -f ``` - ✅ Verify all Continuous Integration (CI) **checks are passing**. - ✅ Reduce changes to the absolute **minimum** required for your bug fix or feature addition. _"It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is."_ — Bruce Lee ## Submitting a Bug Report 🐛 If you spot a problem with YOLOv5 please submit a Bug Report! For us to start investigating a possible problem we need to be able to reproduce it ourselves first. We've created a few short guidelines below to help users provide what we need in order to get started. When asking a question, people will be better able to provide help if you provide **code** that they can easily understand and use to **reproduce** the problem. This is referred to by community members as creating a [minimum reproducible example](https://stackoverflow.com/help/minimal-reproducible-example). Your code that reproduces the problem should be: * ✅ **Minimal** – Use as little code as possible that still produces the same problem * ✅ **Complete** – Provide **all** parts someone else needs to reproduce your problem in the question itself * ✅ **Reproducible** – Test the code you're about to provide to make sure it reproduces the problem In addition to the above requirements, for [Ultralytics](https://ultralytics.com/) to provide assistance your code should be: * ✅ **Current** – Verify that your code is up-to-date with current GitHub [master](https://github.com/ultralytics/yolov5/tree/master), and if necessary `git pull` or `git clone` a new copy to ensure your problem has not already been resolved by previous commits. * ✅ **Unmodified** – Your problem must be reproducible without any modifications to the codebase in this repository. [Ultralytics](https://ultralytics.com/) does not provide support for custom code ⚠️. If you believe your problem meets all of the above criteria, please close this issue and raise a new one using the 🐛 ** Bug Report** [template](https://github.com/ultralytics/yolov5/issues/new/choose) and providing a [minimum reproducible example](https://stackoverflow.com/help/minimal-reproducible-example) to help us better understand and diagnose your problem. ## License By contributing, you agree that your contributions will be licensed under the [GPL-3.0 license](https://choosealicense.com/licenses/gpl-3.0/) ================================================ FILE: Dockerfile ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Start FROM Nvidia PyTorch image https://ngc.nvidia.com/catalog/containers/nvidia:pytorch FROM nvcr.io/nvidia/pytorch:21.10-py3 # Install linux packages RUN apt update && apt install -y zip htop screen libgl1-mesa-glx # Install python dependencies COPY requirements.txt . RUN python -m pip install --upgrade pip RUN pip uninstall -y nvidia-tensorboard nvidia-tensorboard-plugin-dlprof RUN pip install --no-cache -r requirements.txt coremltools onnx gsutil notebook wandb>=0.12.2 RUN pip install --no-cache -U torch torchvision numpy Pillow # RUN pip install --no-cache torch==1.10.0+cu113 torchvision==0.11.1+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html # Create working directory RUN mkdir -p /usr/src/app WORKDIR /usr/src/app # Copy contents COPY . /usr/src/app # Downloads to user config dir ADD https://ultralytics.com/assets/Arial.ttf /root/.config/Ultralytics/ # Set environment variables # ENV HOME=/usr/src/app # Usage Examples ------------------------------------------------------------------------------------------------------- # Build and Push # t=ultralytics/yolov5:latest && sudo docker build -t $t . && sudo docker push $t # Pull and Run # t=ultralytics/yolov5:latest && sudo docker pull $t && sudo docker run -it --ipc=host --gpus all $t # Pull and Run with local directory access # t=ultralytics/yolov5:latest && sudo docker pull $t && sudo docker run -it --ipc=host --gpus all -v "$(pwd)"/datasets:/usr/src/datasets $t # Kill all # sudo docker kill $(sudo docker ps -q) # Kill all image-based # sudo docker kill $(sudo docker ps -qa --filter ancestor=ultralytics/yolov5:latest) # Bash into running container # sudo docker exec -it 5a9b5863d93d bash # Bash into stopped container # id=$(sudo docker ps -qa) && sudo docker start $id && sudo docker exec -it $id bash # Clean up # docker system prune -a --volumes # Update Ubuntu drivers # https://www.maketecheasier.com/install-nvidia-drivers-ubuntu/ # DDP test # python -m torch.distributed.run --nproc_per_node 2 --master_port 1 train.py --epochs 3 # GCP VM from Image # docker.io/ultralytics/yolov5:latest ================================================ FILE: LICENSE ================================================ GNU GENERAL PUBLIC LICENSE Version 3, 29 June 2007 Copyright (C) 2007 Free Software Foundation, Inc. Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. Preamble The GNU General Public License is a free, copyleft license for software and other kinds of works. The licenses for most software and other practical works are designed to take away your freedom to share and change the works. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change all versions of a program--to make sure it remains free software for all its users. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work released this way by its authors. You can apply it to your programs, too. When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for them if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs, and that you know you can do these things. To protect your rights, we need to prevent others from denying you these rights or asking you to surrender the rights. Therefore, you have certain responsibilities if you distribute copies of the software, or if you modify it: responsibilities to respect the freedom of others. For example, if you distribute copies of such a program, whether gratis or for a fee, you must pass on to the recipients the same freedoms that you received. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights. Developers that use the GNU GPL protect your rights with two steps: (1) assert copyright on the software, and (2) offer you this License giving you legal permission to copy, distribute and/or modify it. For the developers' and authors' protection, the GPL clearly explains that there is no warranty for this free software. For both users' and authors' sake, the GPL requires that modified versions be marked as changed, so that their problems will not be attributed erroneously to authors of previous versions. Some devices are designed to deny users access to install or run modified versions of the software inside them, although the manufacturer can do so. This is fundamentally incompatible with the aim of protecting users' freedom to change the software. The systematic pattern of such abuse occurs in the area of products for individuals to use, which is precisely where it is most unacceptable. Therefore, we have designed this version of the GPL to prohibit the practice for those products. If such problems arise substantially in other domains, we stand ready to extend this provision to those domains in future versions of the GPL, as needed to protect the freedom of users. Finally, every program is threatened constantly by software patents. States should not allow patents to restrict development and use of software on general-purpose computers, but in those that do, we wish to avoid the special danger that patents applied to a free program could make it effectively proprietary. To prevent this, the GPL assures that patents cannot be used to render the program non-free. The precise terms and conditions for copying, distribution and modification follow. TERMS AND CONDITIONS 0. Definitions. "This License" refers to version 3 of the GNU General Public License. "Copyright" also means copyright-like laws that apply to other kinds of works, such as semiconductor masks. "The Program" refers to any copyrightable work licensed under this License. Each licensee is addressed as "you". "Licensees" and "recipients" may be individuals or organizations. To "modify" a work means to copy from or adapt all or part of the work in a fashion requiring copyright permission, other than the making of an exact copy. The resulting work is called a "modified version" of the earlier work or a work "based on" the earlier work. A "covered work" means either the unmodified Program or a work based on the Program. To "propagate" a work means to do anything with it that, without permission, would make you directly or secondarily liable for infringement under applicable copyright law, except executing it on a computer or modifying a private copy. Propagation includes copying, distribution (with or without modification), making available to the public, and in some countries other activities as well. To "convey" a work means any kind of propagation that enables other parties to make or receive copies. Mere interaction with a user through a computer network, with no transfer of a copy, is not conveying. An interactive user interface displays "Appropriate Legal Notices" to the extent that it includes a convenient and prominently visible feature that (1) displays an appropriate copyright notice, and (2) tells the user that there is no warranty for the work (except to the extent that warranties are provided), that licensees may convey the work under this License, and how to view a copy of this License. If the interface presents a list of user commands or options, such as a menu, a prominent item in the list meets this criterion. 1. Source Code. The "source code" for a work means the preferred form of the work for making modifications to it. "Object code" means any non-source form of a work. A "Standard Interface" means an interface that either is an official standard defined by a recognized standards body, or, in the case of interfaces specified for a particular programming language, one that is widely used among developers working in that language. The "System Libraries" of an executable work include anything, other than the work as a whole, that (a) is included in the normal form of packaging a Major Component, but which is not part of that Major Component, and (b) serves only to enable use of the work with that Major Component, or to implement a Standard Interface for which an implementation is available to the public in source code form. A "Major Component", in this context, means a major essential component (kernel, window system, and so on) of the specific operating system (if any) on which the executable work runs, or a compiler used to produce the work, or an object code interpreter used to run it. The "Corresponding Source" for a work in object code form means all the source code needed to generate, install, and (for an executable work) run the object code and to modify the work, including scripts to control those activities. However, it does not include the work's System Libraries, or general-purpose tools or generally available free programs which are used unmodified in performing those activities but which are not part of the work. For example, Corresponding Source includes interface definition files associated with source files for the work, and the source code for shared libraries and dynamically linked subprograms that the work is specifically designed to require, such as by intimate data communication or control flow between those subprograms and other parts of the work. The Corresponding Source need not include anything that users can regenerate automatically from other parts of the Corresponding Source. The Corresponding Source for a work in source code form is that same work. 2. Basic Permissions. All rights granted under this License are granted for the term of copyright on the Program, and are irrevocable provided the stated conditions are met. This License explicitly affirms your unlimited permission to run the unmodified Program. The output from running a covered work is covered by this License only if the output, given its content, constitutes a covered work. This License acknowledges your rights of fair use or other equivalent, as provided by copyright law. You may make, run and propagate covered works that you do not convey, without conditions so long as your license otherwise remains in force. You may convey covered works to others for the sole purpose of having them make modifications exclusively for you, or provide you with facilities for running those works, provided that you comply with the terms of this License in conveying all material for which you do not control copyright. Those thus making or running the covered works for you must do so exclusively on your behalf, under your direction and control, on terms that prohibit them from making any copies of your copyrighted material outside their relationship with you. Conveying under any other circumstances is permitted solely under the conditions stated below. Sublicensing is not allowed; section 10 makes it unnecessary. 3. Protecting Users' Legal Rights From Anti-Circumvention Law. No covered work shall be deemed part of an effective technological measure under any applicable law fulfilling obligations under article 11 of the WIPO copyright treaty adopted on 20 December 1996, or similar laws prohibiting or restricting circumvention of such measures. When you convey a covered work, you waive any legal power to forbid circumvention of technological measures to the extent such circumvention is effected by exercising rights under this License with respect to the covered work, and you disclaim any intention to limit operation or modification of the work as a means of enforcing, against the work's users, your or third parties' legal rights to forbid circumvention of technological measures. 4. Conveying Verbatim Copies. You may convey verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice; keep intact all notices stating that this License and any non-permissive terms added in accord with section 7 apply to the code; keep intact all notices of the absence of any warranty; and give all recipients a copy of this License along with the Program. You may charge any price or no price for each copy that you convey, and you may offer support or warranty protection for a fee. 5. Conveying Modified Source Versions. You may convey a work based on the Program, or the modifications to produce it from the Program, in the form of source code under the terms of section 4, provided that you also meet all of these conditions: a) The work must carry prominent notices stating that you modified it, and giving a relevant date. b) The work must carry prominent notices stating that it is released under this License and any conditions added under section 7. This requirement modifies the requirement in section 4 to "keep intact all notices". c) You must license the entire work, as a whole, under this License to anyone who comes into possession of a copy. This License will therefore apply, along with any applicable section 7 additional terms, to the whole of the work, and all its parts, regardless of how they are packaged. This License gives no permission to license the work in any other way, but it does not invalidate such permission if you have separately received it. d) If the work has interactive user interfaces, each must display Appropriate Legal Notices; however, if the Program has interactive interfaces that do not display Appropriate Legal Notices, your work need not make them do so. A compilation of a covered work with other separate and independent works, which are not by their nature extensions of the covered work, and which are not combined with it such as to form a larger program, in or on a volume of a storage or distribution medium, is called an "aggregate" if the compilation and its resulting copyright are not used to limit the access or legal rights of the compilation's users beyond what the individual works permit. Inclusion of a covered work in an aggregate does not cause this License to apply to the other parts of the aggregate. 6. Conveying Non-Source Forms. You may convey a covered work in object code form under the terms of sections 4 and 5, provided that you also convey the machine-readable Corresponding Source under the terms of this License, in one of these ways: a) Convey the object code in, or embodied in, a physical product (including a physical distribution medium), accompanied by the Corresponding Source fixed on a durable physical medium customarily used for software interchange. b) Convey the object code in, or embodied in, a physical product (including a physical distribution medium), accompanied by a written offer, valid for at least three years and valid for as long as you offer spare parts or customer support for that product model, to give anyone who possesses the object code either (1) a copy of the Corresponding Source for all the software in the product that is covered by this License, on a durable physical medium customarily used for software interchange, for a price no more than your reasonable cost of physically performing this conveying of source, or (2) access to copy the Corresponding Source from a network server at no charge. c) Convey individual copies of the object code with a copy of the written offer to provide the Corresponding Source. This alternative is allowed only occasionally and noncommercially, and only if you received the object code with such an offer, in accord with subsection 6b. d) Convey the object code by offering access from a designated place (gratis or for a charge), and offer equivalent access to the Corresponding Source in the same way through the same place at no further charge. You need not require recipients to copy the Corresponding Source along with the object code. If the place to copy the object code is a network server, the Corresponding Source may be on a different server (operated by you or a third party) that supports equivalent copying facilities, provided you maintain clear directions next to the object code saying where to find the Corresponding Source. Regardless of what server hosts the Corresponding Source, you remain obligated to ensure that it is available for as long as needed to satisfy these requirements. e) Convey the object code using peer-to-peer transmission, provided you inform other peers where the object code and Corresponding Source of the work are being offered to the general public at no charge under subsection 6d. A separable portion of the object code, whose source code is excluded from the Corresponding Source as a System Library, need not be included in conveying the object code work. A "User Product" is either (1) a "consumer product", which means any tangible personal property which is normally used for personal, family, or household purposes, or (2) anything designed or sold for incorporation into a dwelling. In determining whether a product is a consumer product, doubtful cases shall be resolved in favor of coverage. For a particular product received by a particular user, "normally used" refers to a typical or common use of that class of product, regardless of the status of the particular user or of the way in which the particular user actually uses, or expects or is expected to use, the product. A product is a consumer product regardless of whether the product has substantial commercial, industrial or non-consumer uses, unless such uses represent the only significant mode of use of the product. "Installation Information" for a User Product means any methods, procedures, authorization keys, or other information required to install and execute modified versions of a covered work in that User Product from a modified version of its Corresponding Source. The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with solely because modification has been made. If you convey an object code work under this section in, or with, or specifically for use in, a User Product, and the conveying occurs as part of a transaction in which the right of possession and use of the User Product is transferred to the recipient in perpetuity or for a fixed term (regardless of how the transaction is characterized), the Corresponding Source conveyed under this section must be accompanied by the Installation Information. But this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM). The requirement to provide Installation Information does not include a requirement to continue to provide support service, warranty, or updates for a work that has been modified or installed by the recipient, or for the User Product in which it has been modified or installed. Access to a network may be denied when the modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network. Corresponding Source conveyed, and Installation Information provided, in accord with this section must be in a format that is publicly documented (and with an implementation available to the public in source code form), and must require no special password or key for unpacking, reading or copying. 7. Additional Terms. "Additional permissions" are terms that supplement the terms of this License by making exceptions from one or more of its conditions. Additional permissions that are applicable to the entire Program shall be treated as though they were included in this License, to the extent that they are valid under applicable law. If additional permissions apply only to part of the Program, that part may be used separately under those permissions, but the entire Program remains governed by this License without regard to the additional permissions. When you convey a copy of a covered work, you may at your option remove any additional permissions from that copy, or from any part of it. (Additional permissions may be written to require their own removal in certain cases when you modify the work.) You may place additional permissions on material, added by you to a covered work, for which you have or can give appropriate copyright permission. Notwithstanding any other provision of this License, for material you add to a covered work, you may (if authorized by the copyright holders of that material) supplement the terms of this License with terms: a) Disclaiming warranty or limiting liability differently from the terms of sections 15 and 16 of this License; or b) Requiring preservation of specified reasonable legal notices or author attributions in that material or in the Appropriate Legal Notices displayed by works containing it; or c) Prohibiting misrepresentation of the origin of that material, or requiring that modified versions of such material be marked in reasonable ways as different from the original version; or d) Limiting the use for publicity purposes of names of licensors or authors of the material; or e) Declining to grant rights under trademark law for use of some trade names, trademarks, or service marks; or f) Requiring indemnification of licensors and authors of that material by anyone who conveys the material (or modified versions of it) with contractual assumptions of liability to the recipient, for any liability that these contractual assumptions directly impose on those licensors and authors. All other non-permissive additional terms are considered "further restrictions" within the meaning of section 10. If the Program as you received it, or any part of it, contains a notice stating that it is governed by this License along with a term that is a further restriction, you may remove that term. If a license document contains a further restriction but permits relicensing or conveying under this License, you may add to a covered work material governed by the terms of that license document, provided that the further restriction does not survive such relicensing or conveying. If you add terms to a covered work in accord with this section, you must place, in the relevant source files, a statement of the additional terms that apply to those files, or a notice indicating where to find the applicable terms. Additional terms, permissive or non-permissive, may be stated in the form of a separately written license, or stated as exceptions; the above requirements apply either way. 8. Termination. You may not propagate or modify a covered work except as expressly provided under this License. Any attempt otherwise to propagate or modify it is void, and will automatically terminate your rights under this License (including any patent licenses granted under the third paragraph of section 11). However, if you cease all violation of this License, then your license from a particular copyright holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and finally terminates your license, and (b) permanently, if the copyright holder fails to notify you of the violation by some reasonable means prior to 60 days after the cessation. Moreover, your license from a particular copyright holder is reinstated permanently if the copyright holder notifies you of the violation by some reasonable means, this is the first time you have received notice of violation of this License (for any work) from that copyright holder, and you cure the violation prior to 30 days after your receipt of the notice. Termination of your rights under this section does not terminate the licenses of parties who have received copies or rights from you under this License. If your rights have been terminated and not permanently reinstated, you do not qualify to receive new licenses for the same material under section 10. 9. Acceptance Not Required for Having Copies. You are not required to accept this License in order to receive or run a copy of the Program. Ancillary propagation of a covered work occurring solely as a consequence of using peer-to-peer transmission to receive a copy likewise does not require acceptance. However, nothing other than this License grants you permission to propagate or modify any covered work. These actions infringe copyright if you do not accept this License. Therefore, by modifying or propagating a covered work, you indicate your acceptance of this License to do so. 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. You are not responsible for enforcing compliance by third parties with this License. An "entity transaction" is a transaction transferring control of an organization, or substantially all assets of one, or subdividing an organization, or merging organizations. If propagation of a covered work results from an entity transaction, each party to that transaction who receives a copy of the work also receives whatever licenses to the work the party's predecessor in interest had or could give under the previous paragraph, plus a right to possession of the Corresponding Source of the work from the predecessor in interest, if the predecessor has it or can get it with reasonable efforts. You may not impose any further restrictions on the exercise of the rights granted or affirmed under this License. For example, you may not impose a license fee, royalty, or other charge for exercise of rights granted under this License, and you may not initiate litigation (including a cross-claim or counterclaim in a lawsuit) alleging that any patent claim is infringed by making, using, selling, offering for sale, or importing the Program or any portion of it. 11. Patents. A "contributor" is a copyright holder who authorizes use under this License of the Program or a work on which the Program is based. The work thus licensed is called the contributor's "contributor version". A contributor's "essential patent claims" are all patent claims owned or controlled by the contributor, whether already acquired or hereafter acquired, that would be infringed by some manner, permitted by this License, of making, using, or selling its contributor version, but do not include claims that would be infringed only as a consequence of further modification of the contributor version. For purposes of this definition, "control" includes the right to grant patent sublicenses in a manner consistent with the requirements of this License. Each contributor grants you a non-exclusive, worldwide, royalty-free patent license under the contributor's essential patent claims, to make, use, sell, offer for sale, import and otherwise run, modify and propagate the contents of its contributor version. In the following three paragraphs, a "patent license" is any express agreement or commitment, however denominated, not to enforce a patent (such as an express permission to practice a patent or covenant not to sue for patent infringement). To "grant" such a patent license to a party means to make such an agreement or commitment not to enforce a patent against the party. If you convey a covered work, knowingly relying on a patent license, and the Corresponding Source of the work is not available for anyone to copy, free of charge and under the terms of this License, through a publicly available network server or other readily accessible means, then you must either (1) cause the Corresponding Source to be so available, or (2) arrange to deprive yourself of the benefit of the patent license for this particular work, or (3) arrange, in a manner consistent with the requirements of this License, to extend the patent license to downstream recipients. "Knowingly relying" means you have actual knowledge that, but for the patent license, your conveying the covered work in a country, or your recipient's use of the covered work in a country, would infringe one or more identifiable patents in that country that you have reason to believe are valid. If, pursuant to or in connection with a single transaction or arrangement, you convey, or propagate by procuring conveyance of, a covered work, and grant a patent license to some of the parties receiving the covered work authorizing them to use, propagate, modify or convey a specific copy of the covered work, then the patent license you grant is automatically extended to all recipients of the covered work and works based on it. A patent license is "discriminatory" if it does not include within the scope of its coverage, prohibits the exercise of, or is conditioned on the non-exercise of one or more of the rights that are specifically granted under this License. You may not convey a covered work if you are a party to an arrangement with a third party that is in the business of distributing software, under which you make payment to the third party based on the extent of your activity of conveying the work, and under which the third party grants, to any of the parties who would receive the covered work from you, a discriminatory patent license (a) in connection with copies of the covered work conveyed by you (or copies made from those copies), or (b) primarily for and in connection with specific products or compilations that contain the covered work, unless you entered into that arrangement, or that patent license was granted, prior to 28 March 2007. Nothing in this License shall be construed as excluding or limiting any implied license or other defenses to infringement that may otherwise be available to you under applicable patent law. 12. No Surrender of Others' Freedom. If conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot convey a covered work so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not convey it at all. For example, if you agree to terms that obligate you to collect a royalty for further conveying from those to whom you convey the Program, the only way you could satisfy both those terms and this License would be to refrain entirely from conveying the Program. 13. Use with the GNU Affero General Public License. Notwithstanding any other provision of this License, you have permission to link or combine any covered work with a work licensed under version 3 of the GNU Affero General Public License into a single combined work, and to convey the resulting work. The terms of this License will continue to apply to the part which is the covered work, but the special requirements of the GNU Affero General Public License, section 13, concerning interaction through a network will apply to the combination as such. 14. Revised Versions of this License. The Free Software Foundation may publish revised and/or new versions of the GNU General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies that a certain numbered version of the GNU General Public License "or any later version" applies to it, you have the option of following the terms and conditions either of that numbered version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of the GNU General Public License, you may choose any version ever published by the Free Software Foundation. If the Program specifies that a proxy can decide which future versions of the GNU General Public License can be used, that proxy's public statement of acceptance of a version permanently authorizes you to choose that version for the Program. Later license versions may give you additional or different permissions. However, no additional obligations are imposed on any author or copyright holder as a result of your choosing to follow a later version. 15. Disclaimer of Warranty. THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 16. Limitation of Liability. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. 17. Interpretation of Sections 15 and 16. If the disclaimer of warranty and limitation of liability provided above cannot be given local legal effect according to their terms, reviewing courts shall apply local law that most closely approximates an absolute waiver of all civil liability in connection with the Program, unless a warranty or assumption of liability accompanies a copy of the Program in return for a fee. END OF TERMS AND CONDITIONS How to Apply These Terms to Your New Programs If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms. To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively state the exclusion of warranty; and each file should have at least the "copyright" line and a pointer to where the full notice is found. Copyright (C) This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see . Also add information on how to contact you by electronic and paper mail. If the program does terminal interaction, make it output a short notice like this when it starts in an interactive mode: Copyright (C) This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details. The hypothetical commands `show w' and `show c' should show the appropriate parts of the General Public License. Of course, your program's commands might be different; for a GUI interface, you would use an "about box". You should also get your employer (if you work as a programmer) or school, if any, to sign a "copyright disclaimer" for the program, if necessary. For more information on this, and how to apply and follow the GNU GPL, see . The GNU General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Lesser General Public License instead of this License. But first, please read . ================================================ FILE: README.md ================================================ ## 仓库说明： 本仓库是针对基于EASY-EAI-Nano(RV1126)从PC端模型训练、模型单步测试、pytorch模型转换为onnx模型的流程说明，并以口罩检测为例子说明。而模型如何部署到硬件主板上，完整的在线文档教程可以查看以下在线文档的链接： ## 环境说明： python version >= 3.6 pytorch version >= 1.7 onnx verison >= 1.11 ## 准备数据 口罩检测数据百度链接：https://pan.baidu.com/s/1vtxWurn1Mqu-wJ017eaQrw 提取码：6666 数据集解压后(脚本在数据集里面)，执行以下脚本生成train.txt和valid.txt： ```python python list_dataset_file.py ``` ## 训练模型 训练一个口罩检测模型，需要修改"data/mask.yaml"里面的train.txt和valid.txt的路径。训练脚本如下所示： ```python python train.py --data mask.yaml --cfg yolov5s.yaml --weights "" --batch-size 64 yolov5m 40 yolov5l 24 yolov5x 16 ``` 训练完成后会在 ## 模型预测 测试训练好的模型： ```python python detect.py --source data/images --weights ./runs/train/exp/weights/best.pt --conf 0.5 ``` 测试结果会在"runs/detect"生成： ## 模型导出 执行以下指令把pt模型转换为onnx模型，同时会生成best.anchors.txt： ```python python export.py --include onnx --rknpu RV1126 --weights ./runs/train/exp/weights/best.pt ``` ### EASY-EAI-Nano基于NPU运行速度测试(单位:ms)： | 模型(640x640输入)         | EASY-EAI-Nano(RV1126) | | :---------------------- | :-----------------------------------------: | | yolov5s int8量化 |   52    | | yolov5m int8量化 |   93    | ## 参考库： https://github.com/ultralytics/yolov5 https://github.com/soloIife/yolov5_for_rknn ## 技术交流群： QQ群：810456486 ================================================ FILE: data/Argoverse.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Argoverse-HD dataset (ring-front-center camera) http://www.cs.cmu.edu/~mengtial/proj/streaming/ # Example usage: python train.py --data Argoverse.yaml # parent # ├── yolov5 # └── datasets # └── Argoverse ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] path: ../datasets/Argoverse # dataset root dir train: Argoverse-1.1/images/train/ # train images (relative to 'path') 39384 images val: Argoverse-1.1/images/val/ # val images (relative to 'path') 15062 images test: Argoverse-1.1/images/test/ # test images (optional) https://eval.ai/web/challenges/challenge-page/800/overview # Classes nc: 8 # number of classes names: ['person', 'bicycle', 'car', 'motorcycle', 'bus', 'truck', 'traffic_light', 'stop_sign'] # class names # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | import json from tqdm import tqdm from utils.general import download, Path def argoverse2yolo(set): labels = {} a = json.load(open(set, "rb")) for annot in tqdm(a['annotations'], desc=f"Converting {set} to YOLOv5 format..."): img_id = annot['image_id'] img_name = a['images'][img_id]['name'] img_label_name = img_name[:-3] + "txt" cls = annot['category_id'] # instance class id x_center, y_center, width, height = annot['bbox'] x_center = (x_center + width / 2) / 1920.0 # offset and scale y_center = (y_center + height / 2) / 1200.0 # offset and scale width /= 1920.0 # scale height /= 1200.0 # scale img_dir = set.parents[2] / 'Argoverse-1.1' / 'labels' / a['seq_dirs'][a['images'][annot['image_id']]['sid']] if not img_dir.exists(): img_dir.mkdir(parents=True, exist_ok=True) k = str(img_dir / img_label_name) if k not in labels: labels[k] = [] labels[k].append(f"{cls} {x_center} {y_center} {width} {height}\n") for k in labels: with open(k, "w") as f: f.writelines(labels[k]) # Download dir = Path('../datasets/Argoverse') # dataset root dir urls = ['https://argoverse-hd.s3.us-east-2.amazonaws.com/Argoverse-HD-Full.zip'] download(urls, dir=dir, delete=False) # Convert annotations_dir = 'Argoverse-HD/annotations/' (dir / 'Argoverse-1.1' / 'tracking').rename(dir / 'Argoverse-1.1' / 'images') # rename 'tracking' to 'images' for d in "train.json", "val.json": argoverse2yolo(dir / annotations_dir / d) # convert VisDrone annotations to YOLO labels ================================================ FILE: data/GlobalWheat2020.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Global Wheat 2020 dataset http://www.global-wheat.com/ # Example usage: python train.py --data GlobalWheat2020.yaml # parent # ├── yolov5 # └── datasets # └── GlobalWheat2020 ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] path: ../datasets/GlobalWheat2020 # dataset root dir train: # train images (relative to 'path') 3422 images - images/arvalis_1 - images/arvalis_2 - images/arvalis_3 - images/ethz_1 - images/rres_1 - images/inrae_1 - images/usask_1 val: # val images (relative to 'path') 748 images (WARNING: train set contains ethz_1) - images/ethz_1 test: # test images (optional) 1276 images - images/utokyo_1 - images/utokyo_2 - images/nau_1 - images/uq_1 # Classes nc: 1 # number of classes names: ['wheat_head'] # class names # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | from utils.general import download, Path # Download dir = Path(yaml['path']) # dataset root dir urls = ['https://zenodo.org/record/4298502/files/global-wheat-codalab-official.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/GlobalWheat2020_labels.zip'] download(urls, dir=dir) # Make Directories for p in 'annotations', 'images', 'labels': (dir / p).mkdir(parents=True, exist_ok=True) # Move for p in 'arvalis_1', 'arvalis_2', 'arvalis_3', 'ethz_1', 'rres_1', 'inrae_1', 'usask_1', \ 'utokyo_1', 'utokyo_2', 'nau_1', 'uq_1': (dir / p).rename(dir / 'images' / p) # move to /images f = (dir / p).with_suffix('.json') # json file if f.exists(): f.rename((dir / 'annotations' / p).with_suffix('.json')) # move to /annotations ================================================ FILE: data/Objects365.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Objects365 dataset https://www.objects365.org/ # Example usage: python train.py --data Objects365.yaml # parent # ├── yolov5 # └── datasets # └── Objects365 ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] path: ../datasets/Objects365 # dataset root dir train: images/train # train images (relative to 'path') 1742289 images val: images/val # val images (relative to 'path') 80000 images test: # test images (optional) # Classes nc: 365 # number of classes names: ['Person', 'Sneakers', 'Chair', 'Other Shoes', 'Hat', 'Car', 'Lamp', 'Glasses', 'Bottle', 'Desk', 'Cup', 'Street Lights', 'Cabinet/shelf', 'Handbag/Satchel', 'Bracelet', 'Plate', 'Picture/Frame', 'Helmet', 'Book', 'Gloves', 'Storage box', 'Boat', 'Leather Shoes', 'Flower', 'Bench', 'Potted Plant', 'Bowl/Basin', 'Flag', 'Pillow', 'Boots', 'Vase', 'Microphone', 'Necklace', 'Ring', 'SUV', 'Wine Glass', 'Belt', 'Monitor/TV', 'Backpack', 'Umbrella', 'Traffic Light', 'Speaker', 'Watch', 'Tie', 'Trash bin Can', 'Slippers', 'Bicycle', 'Stool', 'Barrel/bucket', 'Van', 'Couch', 'Sandals', 'Basket', 'Drum', 'Pen/Pencil', 'Bus', 'Wild Bird', 'High Heels', 'Motorcycle', 'Guitar', 'Carpet', 'Cell Phone', 'Bread', 'Camera', 'Canned', 'Truck', 'Traffic cone', 'Cymbal', 'Lifesaver', 'Towel', 'Stuffed Toy', 'Candle', 'Sailboat', 'Laptop', 'Awning', 'Bed', 'Faucet', 'Tent', 'Horse', 'Mirror', 'Power outlet', 'Sink', 'Apple', 'Air Conditioner', 'Knife', 'Hockey Stick', 'Paddle', 'Pickup Truck', 'Fork', 'Traffic Sign', 'Balloon', 'Tripod', 'Dog', 'Spoon', 'Clock', 'Pot', 'Cow', 'Cake', 'Dinning Table', 'Sheep', 'Hanger', 'Blackboard/Whiteboard', 'Napkin', 'Other Fish', 'Orange/Tangerine', 'Toiletry', 'Keyboard', 'Tomato', 'Lantern', 'Machinery Vehicle', 'Fan', 'Green Vegetables', 'Banana', 'Baseball Glove', 'Airplane', 'Mouse', 'Train', 'Pumpkin', 'Soccer', 'Skiboard', 'Luggage', 'Nightstand', 'Tea pot', 'Telephone', 'Trolley', 'Head Phone', 'Sports Car', 'Stop Sign', 'Dessert', 'Scooter', 'Stroller', 'Crane', 'Remote', 'Refrigerator', 'Oven', 'Lemon', 'Duck', 'Baseball Bat', 'Surveillance Camera', 'Cat', 'Jug', 'Broccoli', 'Piano', 'Pizza', 'Elephant', 'Skateboard', 'Surfboard', 'Gun', 'Skating and Skiing shoes', 'Gas stove', 'Donut', 'Bow Tie', 'Carrot', 'Toilet', 'Kite', 'Strawberry', 'Other Balls', 'Shovel', 'Pepper', 'Computer Box', 'Toilet Paper', 'Cleaning Products', 'Chopsticks', 'Microwave', 'Pigeon', 'Baseball', 'Cutting/chopping Board', 'Coffee Table', 'Side Table', 'Scissors', 'Marker', 'Pie', 'Ladder', 'Snowboard', 'Cookies', 'Radiator', 'Fire Hydrant', 'Basketball', 'Zebra', 'Grape', 'Giraffe', 'Potato', 'Sausage', 'Tricycle', 'Violin', 'Egg', 'Fire Extinguisher', 'Candy', 'Fire Truck', 'Billiards', 'Converter', 'Bathtub', 'Wheelchair', 'Golf Club', 'Briefcase', 'Cucumber', 'Cigar/Cigarette', 'Paint Brush', 'Pear', 'Heavy Truck', 'Hamburger', 'Extractor', 'Extension Cord', 'Tong', 'Tennis Racket', 'Folder', 'American Football', 'earphone', 'Mask', 'Kettle', 'Tennis', 'Ship', 'Swing', 'Coffee Machine', 'Slide', 'Carriage', 'Onion', 'Green beans', 'Projector', 'Frisbee', 'Washing Machine/Drying Machine', 'Chicken', 'Printer', 'Watermelon', 'Saxophone', 'Tissue', 'Toothbrush', 'Ice cream', 'Hot-air balloon', 'Cello', 'French Fries', 'Scale', 'Trophy', 'Cabbage', 'Hot dog', 'Blender', 'Peach', 'Rice', 'Wallet/Purse', 'Volleyball', 'Deer', 'Goose', 'Tape', 'Tablet', 'Cosmetics', 'Trumpet', 'Pineapple', 'Golf Ball', 'Ambulance', 'Parking meter', 'Mango', 'Key', 'Hurdle', 'Fishing Rod', 'Medal', 'Flute', 'Brush', 'Penguin', 'Megaphone', 'Corn', 'Lettuce', 'Garlic', 'Swan', 'Helicopter', 'Green Onion', 'Sandwich', 'Nuts', 'Speed Limit Sign', 'Induction Cooker', 'Broom', 'Trombone', 'Plum', 'Rickshaw', 'Goldfish', 'Kiwi fruit', 'Router/modem', 'Poker Card', 'Toaster', 'Shrimp', 'Sushi', 'Cheese', 'Notepaper', 'Cherry', 'Pliers', 'CD', 'Pasta', 'Hammer', 'Cue', 'Avocado', 'Hamimelon', 'Flask', 'Mushroom', 'Screwdriver', 'Soap', 'Recorder', 'Bear', 'Eggplant', 'Board Eraser', 'Coconut', 'Tape Measure/Ruler', 'Pig', 'Showerhead', 'Globe', 'Chips', 'Steak', 'Crosswalk Sign', 'Stapler', 'Camel', 'Formula 1', 'Pomegranate', 'Dishwasher', 'Crab', 'Hoverboard', 'Meat ball', 'Rice Cooker', 'Tuba', 'Calculator', 'Papaya', 'Antelope', 'Parrot', 'Seal', 'Butterfly', 'Dumbbell', 'Donkey', 'Lion', 'Urinal', 'Dolphin', 'Electric Drill', 'Hair Dryer', 'Egg tart', 'Jellyfish', 'Treadmill', 'Lighter', 'Grapefruit', 'Game board', 'Mop', 'Radish', 'Baozi', 'Target', 'French', 'Spring Rolls', 'Monkey', 'Rabbit', 'Pencil Case', 'Yak', 'Red Cabbage', 'Binoculars', 'Asparagus', 'Barbell', 'Scallop', 'Noddles', 'Comb', 'Dumpling', 'Oyster', 'Table Tennis paddle', 'Cosmetics Brush/Eyeliner Pencil', 'Chainsaw', 'Eraser', 'Lobster', 'Durian', 'Okra', 'Lipstick', 'Cosmetics Mirror', 'Curling', 'Table Tennis'] # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | from pycocotools.coco import COCO from tqdm import tqdm from utils.general import Path, download, np, xyxy2xywhn # Make Directories dir = Path(yaml['path']) # dataset root dir for p in 'images', 'labels': (dir / p).mkdir(parents=True, exist_ok=True) for q in 'train', 'val': (dir / p / q).mkdir(parents=True, exist_ok=True) # Train, Val Splits for split, patches in [('train', 50 + 1), ('val', 43 + 1)]: print(f"Processing {split} in {patches} patches ...") images, labels = dir / 'images' / split, dir / 'labels' / split # Download url = f"https://dorc.ks3-cn-beijing.ksyun.com/data-set/2020Objects365%E6%95%B0%E6%8D%AE%E9%9B%86/{split}/" if split == 'train': download([f'{url}zhiyuan_objv2_{split}.tar.gz'], dir=dir, delete=False) # annotations json download([f'{url}patch{i}.tar.gz' for i in range(patches)], dir=images, curl=True, delete=False, threads=8) elif split == 'val': download([f'{url}zhiyuan_objv2_{split}.json'], dir=dir, delete=False) # annotations json download([f'{url}images/v1/patch{i}.tar.gz' for i in range(15 + 1)], dir=images, curl=True, delete=False, threads=8) download([f'{url}images/v2/patch{i}.tar.gz' for i in range(16, patches)], dir=images, curl=True, delete=False, threads=8) # Move for f in tqdm(images.rglob('*.jpg'), desc=f'Moving {split} images'): f.rename(images / f.name) # move to /images/{split} # Labels coco = COCO(dir / f'zhiyuan_objv2_{split}.json') names = [x["name"] for x in coco.loadCats(coco.getCatIds())] for cid, cat in enumerate(names): catIds = coco.getCatIds(catNms=[cat]) imgIds = coco.getImgIds(catIds=catIds) for im in tqdm(coco.loadImgs(imgIds), desc=f'Class {cid + 1}/{len(names)} {cat}'): width, height = im["width"], im["height"] path = Path(im["file_name"]) # image filename try: with open(labels / path.with_suffix('.txt').name, 'a') as file: annIds = coco.getAnnIds(imgIds=im["id"], catIds=catIds, iscrowd=None) for a in coco.loadAnns(annIds): x, y, w, h = a['bbox'] # bounding box in xywh (xy top-left corner) xyxy = np.array([x, y, x + w, y + h])[None] # pixels(1,4) x, y, w, h = xyxy2xywhn(xyxy, w=width, h=height, clip=True)[0] # normalized and clipped file.write(f"{cid} {x:.5f} {y:.5f} {w:.5f} {h:.5f}\n") except Exception as e: print(e) ================================================ FILE: data/SKU-110K.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # SKU-110K retail items dataset https://github.com/eg4000/SKU110K_CVPR19 # Example usage: python train.py --data SKU-110K.yaml # parent # ├── yolov5 # └── datasets # └── SKU-110K ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] path: ../datasets/SKU-110K # dataset root dir train: train.txt # train images (relative to 'path') 8219 images val: val.txt # val images (relative to 'path') 588 images test: test.txt # test images (optional) 2936 images # Classes nc: 1 # number of classes names: ['object'] # class names # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | import shutil from tqdm import tqdm from utils.general import np, pd, Path, download, xyxy2xywh # Download dir = Path(yaml['path']) # dataset root dir parent = Path(dir.parent) # download dir urls = ['http://trax-geometry.s3.amazonaws.com/cvpr_challenge/SKU110K_fixed.tar.gz'] download(urls, dir=parent, delete=False) # Rename directories if dir.exists(): shutil.rmtree(dir) (parent / 'SKU110K_fixed').rename(dir) # rename dir (dir / 'labels').mkdir(parents=True, exist_ok=True) # create labels dir # Convert labels names = 'image', 'x1', 'y1', 'x2', 'y2', 'class', 'image_width', 'image_height' # column names for d in 'annotations_train.csv', 'annotations_val.csv', 'annotations_test.csv': x = pd.read_csv(dir / 'annotations' / d, names=names).values # annotations images, unique_images = x[:, 0], np.unique(x[:, 0]) with open((dir / d).with_suffix('.txt').__str__().replace('annotations_', ''), 'w') as f: f.writelines(f'./images/{s}\n' for s in unique_images) for im in tqdm(unique_images, desc=f'Converting {dir / d}'): cls = 0 # single-class dataset with open((dir / 'labels' / im).with_suffix('.txt'), 'a') as f: for r in x[images == im]: w, h = r[6], r[7] # image width, height xywh = xyxy2xywh(np.array([[r[1] / w, r[2] / h, r[3] / w, r[4] / h]]))[0] # instance f.write(f"{cls} {xywh[0]:.5f} {xywh[1]:.5f} {xywh[2]:.5f} {xywh[3]:.5f}\n") # write label ================================================ FILE: data/VisDrone.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # VisDrone2019-DET dataset https://github.com/VisDrone/VisDrone-Dataset # Example usage: python train.py --data VisDrone.yaml # parent # ├── yolov5 # └── datasets # └── VisDrone ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] path: ../datasets/VisDrone # dataset root dir train: VisDrone2019-DET-train/images # train images (relative to 'path') 6471 images val: VisDrone2019-DET-val/images # val images (relative to 'path') 548 images test: VisDrone2019-DET-test-dev/images # test images (optional) 1610 images # Classes nc: 10 # number of classes names: ['pedestrian', 'people', 'bicycle', 'car', 'van', 'truck', 'tricycle', 'awning-tricycle', 'bus', 'motor'] # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | from utils.general import download, os, Path def visdrone2yolo(dir): from PIL import Image from tqdm import tqdm def convert_box(size, box): # Convert VisDrone box to YOLO xywh box dw = 1. / size[0] dh = 1. / size[1] return (box[0] + box[2] / 2) * dw, (box[1] + box[3] / 2) * dh, box[2] * dw, box[3] * dh (dir / 'labels').mkdir(parents=True, exist_ok=True) # make labels directory pbar = tqdm((dir / 'annotations').glob('*.txt'), desc=f'Converting {dir}') for f in pbar: img_size = Image.open((dir / 'images' / f.name).with_suffix('.jpg')).size lines = [] with open(f, 'r') as file: # read annotation.txt for row in [x.split(',') for x in file.read().strip().splitlines()]: if row[4] == '0': # VisDrone 'ignored regions' class 0 continue cls = int(row[5]) - 1 box = convert_box(img_size, tuple(map(int, row[:4]))) lines.append(f"{cls} {' '.join(f'{x:.6f}' for x in box)}\n") with open(str(f).replace(os.sep + 'annotations' + os.sep, os.sep + 'labels' + os.sep), 'w') as fl: fl.writelines(lines) # write label.txt # Download dir = Path(yaml['path']) # dataset root dir urls = ['https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-train.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-val.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-test-dev.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-test-challenge.zip'] download(urls, dir=dir) # Convert for d in 'VisDrone2019-DET-train', 'VisDrone2019-DET-val', 'VisDrone2019-DET-test-dev': visdrone2yolo(dir / d) # convert VisDrone annotations to YOLO labels ================================================ FILE: data/coco.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # COCO 2017 dataset http://cocodataset.org # Example usage: python train.py --data coco.yaml # parent # ├── yolov5 # └── datasets # └── coco ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] train: D:/ai_project/coco/coco2017labels/coco/train2017.txt # 118287 images val: D:/ai_project/coco/coco2017labels/coco/val2017.txt # 5000 images test: D:/ai_project/coco/coco2017labels/coco/test-dev2017.txt # 20288 of 40670 images, submit to https://competitions.codalab.org/competitions/20794 # Classes nc: 80 # number of classes names: ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light', 'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee', 'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard', 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch', 'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear', 'hair drier', 'toothbrush'] # class names # Download script/URL (optional) ================================================ FILE: data/coco128.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # COCO128 dataset https://www.kaggle.com/ultralytics/coco128 (first 128 images from COCO train2017) # Example usage: python train.py --data coco128.yaml # parent # ├── yolov5 # └── datasets # └── coco128 ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] path: ../datasets/coco128 # dataset root dir train: images/train2017 # train images (relative to 'path') 128 images val: images/train2017 # val images (relative to 'path') 128 images test: # test images (optional) # Classes nc: 80 # number of classes names: ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light', 'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee', 'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard', 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch', 'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear', 'hair drier', 'toothbrush'] # class names # Download script/URL (optional) download: https://ultralytics.com/assets/coco128.zip ================================================ FILE: data/hyps/hyp.finetune.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Hyperparameters for VOC finetuning # python train.py --batch 64 --weights yolov5m.pt --data VOC.yaml --img 512 --epochs 50 # See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials # Hyperparameter Evolution Results # Generations: 306 # P R mAP.5 mAP.5:.95 box obj cls # Metrics: 0.6 0.936 0.896 0.684 0.0115 0.00805 0.00146 lr0: 0.0032 lrf: 0.12 momentum: 0.843 weight_decay: 0.00036 warmup_epochs: 2.0 warmup_momentum: 0.5 warmup_bias_lr: 0.05 box: 0.0296 cls: 0.243 cls_pw: 0.631 obj: 0.301 obj_pw: 0.911 iou_t: 0.2 anchor_t: 2.91 # anchors: 3.63 fl_gamma: 0.0 hsv_h: 0.0138 hsv_s: 0.664 hsv_v: 0.464 degrees: 0.373 translate: 0.245 scale: 0.898 shear: 0.602 perspective: 0.0 flipud: 0.00856 fliplr: 0.5 mosaic: 1.0 mixup: 0.243 copy_paste: 0.0 ================================================ FILE: data/hyps/hyp.finetune_objects365.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license lr0: 0.00258 lrf: 0.17 momentum: 0.779 weight_decay: 0.00058 warmup_epochs: 1.33 warmup_momentum: 0.86 warmup_bias_lr: 0.0711 box: 0.0539 cls: 0.299 cls_pw: 0.825 obj: 0.632 obj_pw: 1.0 iou_t: 0.2 anchor_t: 3.44 anchors: 3.2 fl_gamma: 0.0 hsv_h: 0.0188 hsv_s: 0.704 hsv_v: 0.36 degrees: 0.0 translate: 0.0902 scale: 0.491 shear: 0.0 perspective: 0.0 flipud: 0.0 fliplr: 0.5 mosaic: 1.0 mixup: 0.0 copy_paste: 0.0 ================================================ FILE: data/hyps/hyp.scratch-high.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Hyperparameters for high-augmentation COCO training from scratch # python train.py --batch 32 --cfg yolov5m6.yaml --weights '' --data coco.yaml --img 1280 --epochs 300 # See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials lr0: 0.01 # initial learning rate (SGD=1E-2, Adam=1E-3) lrf: 0.2 # final OneCycleLR learning rate (lr0 * lrf) momentum: 0.937 # SGD momentum/Adam beta1 weight_decay: 0.0005 # optimizer weight decay 5e-4 warmup_epochs: 3.0 # warmup epochs (fractions ok) warmup_momentum: 0.8 # warmup initial momentum warmup_bias_lr: 0.1 # warmup initial bias lr box: 0.05 # box loss gain cls: 0.3 # cls loss gain cls_pw: 1.0 # cls BCELoss positive_weight obj: 0.7 # obj loss gain (scale with pixels) obj_pw: 1.0 # obj BCELoss positive_weight iou_t: 0.20 # IoU training threshold anchor_t: 4.0 # anchor-multiple threshold # anchors: 3 # anchors per output layer (0 to ignore) fl_gamma: 0.0 # focal loss gamma (efficientDet default gamma=1.5) hsv_h: 0.015 # image HSV-Hue augmentation (fraction) hsv_s: 0.7 # image HSV-Saturation augmentation (fraction) hsv_v: 0.4 # image HSV-Value augmentation (fraction) degrees: 0.0 # image rotation (+/- deg) translate: 0.1 # image translation (+/- fraction) scale: 0.9 # image scale (+/- gain) shear: 0.0 # image shear (+/- deg) perspective: 0.0 # image perspective (+/- fraction), range 0-0.001 flipud: 0.0 # image flip up-down (probability) fliplr: 0.5 # image flip left-right (probability) mosaic: 1.0 # image mosaic (probability) mixup: 0.1 # image mixup (probability) copy_paste: 0.1 # segment copy-paste (probability) ================================================ FILE: data/hyps/hyp.scratch-low.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Hyperparameters for low-augmentation COCO training from scratch # python train.py --batch 64 --cfg yolov5n6.yaml --weights '' --data coco.yaml --img 640 --epochs 300 --linear # See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials lr0: 0.01 # initial learning rate (SGD=1E-2, Adam=1E-3) lrf: 0.01 # final OneCycleLR learning rate (lr0 * lrf) momentum: 0.937 # SGD momentum/Adam beta1 weight_decay: 0.0005 # optimizer weight decay 5e-4 warmup_epochs: 3.0 # warmup epochs (fractions ok) warmup_momentum: 0.8 # warmup initial momentum warmup_bias_lr: 0.1 # warmup initial bias lr box: 0.05 # box loss gain cls: 0.5 # cls loss gain cls_pw: 1.0 # cls BCELoss positive_weight obj: 1.0 # obj loss gain (scale with pixels) obj_pw: 1.0 # obj BCELoss positive_weight iou_t: 0.20 # IoU training threshold anchor_t: 4.0 # anchor-multiple threshold # anchors: 3 # anchors per output layer (0 to ignore) fl_gamma: 0.0 # focal loss gamma (efficientDet default gamma=1.5) hsv_h: 0.015 # image HSV-Hue augmentation (fraction) hsv_s: 0.7 # image HSV-Saturation augmentation (fraction) hsv_v: 0.4 # image HSV-Value augmentation (fraction) degrees: 0.0 # image rotation (+/- deg) translate: 0.1 # image translation (+/- fraction) scale: 0.5 # image scale (+/- gain) shear: 0.0 # image shear (+/- deg) perspective: 0.0 # image perspective (+/- fraction), range 0-0.001 flipud: 0.0 # image flip up-down (probability) fliplr: 0.5 # image flip left-right (probability) mosaic: 1.0 # image mosaic (probability) mixup: 0.0 # image mixup (probability) copy_paste: 0.0 # segment copy-paste (probability) ================================================ FILE: data/hyps/hyp.scratch-med.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Hyperparameters for medium-augmentation COCO training from scratch # python train.py --batch 32 --cfg yolov5m6.yaml --weights '' --data coco.yaml --img 1280 --epochs 300 # See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials lr0: 0.01 # initial learning rate (SGD=1E-2, Adam=1E-3) lrf: 0.1 # final OneCycleLR learning rate (lr0 * lrf) momentum: 0.937 # SGD momentum/Adam beta1 weight_decay: 0.0005 # optimizer weight decay 5e-4 warmup_epochs: 3.0 # warmup epochs (fractions ok) warmup_momentum: 0.8 # warmup initial momentum warmup_bias_lr: 0.1 # warmup initial bias lr box: 0.05 # box loss gain cls: 0.3 # cls loss gain cls_pw: 1.0 # cls BCELoss positive_weight obj: 0.7 # obj loss gain (scale with pixels) obj_pw: 1.0 # obj BCELoss positive_weight iou_t: 0.20 # IoU training threshold anchor_t: 4.0 # anchor-multiple threshold # anchors: 3 # anchors per output layer (0 to ignore) fl_gamma: 0.0 # focal loss gamma (efficientDet default gamma=1.5) hsv_h: 0.015 # image HSV-Hue augmentation (fraction) hsv_s: 0.7 # image HSV-Saturation augmentation (fraction) hsv_v: 0.4 # image HSV-Value augmentation (fraction) degrees: 0.0 # image rotation (+/- deg) translate: 0.1 # image translation (+/- fraction) scale: 0.9 # image scale (+/- gain) shear: 0.0 # image shear (+/- deg) perspective: 0.0 # image perspective (+/- fraction), range 0-0.001 flipud: 0.0 # image flip up-down (probability) fliplr: 0.5 # image flip left-right (probability) mosaic: 1.0 # image mosaic (probability) mixup: 0.1 # image mixup (probability) copy_paste: 0.0 # segment copy-paste (probability) ================================================ FILE: data/hyps/hyp.scratch.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Hyperparameters for COCO training from scratch # python train.py --batch 40 --cfg yolov5m.yaml --weights '' --data coco.yaml --img 640 --epochs 300 # See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials lr0: 0.01 # initial learning rate (SGD=1E-2, Adam=1E-3) lrf: 0.1 # final OneCycleLR learning rate (lr0 * lrf) momentum: 0.937 # SGD momentum/Adam beta1 weight_decay: 0.0005 # optimizer weight decay 5e-4 warmup_epochs: 3.0 # warmup epochs (fractions ok) warmup_momentum: 0.8 # warmup initial momentum warmup_bias_lr: 0.1 # warmup initial bias lr box: 0.05 # box loss gain cls: 0.5 # cls loss gain cls_pw: 1.0 # cls BCELoss positive_weight obj: 1.0 # obj loss gain (scale with pixels) obj_pw: 1.0 # obj BCELoss positive_weight iou_t: 0.20 # IoU training threshold anchor_t: 4.0 # anchor-multiple threshold # anchors: 3 # anchors per output layer (0 to ignore) fl_gamma: 0.0 # focal loss gamma (efficientDet default gamma=1.5) hsv_h: 0.015 # image HSV-Hue augmentation (fraction) hsv_s: 0.7 # image HSV-Saturation augmentation (fraction) hsv_v: 0.4 # image HSV-Value augmentation (fraction) degrees: 0.0 # image rotation (+/- deg) translate: 0.1 # image translation (+/- fraction) scale: 0.5 # image scale (+/- gain) shear: 0.0 # image shear (+/- deg) perspective: 0.0 # image perspective (+/- fraction), range 0-0.001 flipud: 0.0 # image flip up-down (probability) fliplr: 0.5 # image flip left-right (probability) mosaic: 1.0 # image mosaic (probability) mixup: 0.0 # image mixup (probability) copy_paste: 0.0 # segment copy-paste (probability) ================================================ FILE: data/mask.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # COCO 2017 dataset http://cocodataset.org # Example usage: python train.py --data mask.yaml # parent # ├── yolov5 # └── datasets # └── coco ← downloads here # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] train: E:/dataset/mask/train.txt val: E:/dataset/mask/valid.txt test: E:/dataset/mask/valid.txt # Classes nc: 2 # number of classes names: ['head', 'mask'] # class names # Download script/URL (optional) ================================================ FILE: data/scripts/download_weights.sh ================================================ #!/bin/bash # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Download latest models from https://github.com/ultralytics/yolov5/releases # Example usage: bash path/to/download_weights.sh # parent # └── yolov5 # ├── yolov5s.pt ← downloads here # ├── yolov5m.pt # └── ... python - <= cls >= 0, f'incorrect class index {cls}' # Write YOLO label if id not in shapes: shapes[id] = Image.open(file).size box = xyxy2xywhn(box[None].astype(np.float), w=shapes[id][0], h=shapes[id][1], clip=True) with open((labels / id).with_suffix('.txt'), 'a') as f: f.write(f"{cls} {' '.join(f'{x:.6f}' for x in box[0])}\n") # write label.txt except Exception as e: print(f'WARNING: skipping one label for {file}: {e}') # Download manually from https://challenge.xviewdataset.org dir = Path(yaml['path']) # dataset root dir # urls = ['https://d307kc0mrhucc3.cloudfront.net/train_labels.zip', # train labels # 'https://d307kc0mrhucc3.cloudfront.net/train_images.zip', # 15G, 847 train images # 'https://d307kc0mrhucc3.cloudfront.net/val_images.zip'] # 5G, 282 val images (no labels) # download(urls, dir=dir, delete=False) # Convert labels convert_labels(dir / 'xView_train.geojson') # Move images images = Path(dir / 'images') images.mkdir(parents=True, exist_ok=True) Path(dir / 'train_images').rename(dir / 'images' / 'train') Path(dir / 'val_images').rename(dir / 'images' / 'val') # Split autosplit(dir / 'images' / 'train') ================================================ FILE: detect.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Run inference on images, videos, directories, streams, etc. Usage: $ python path/to/detect.py --weights yolov5s.pt --source 0 # webcam img.jpg # image vid.mp4 # video path/ # directory path/*.jpg # glob 'https://youtu.be/Zgi9g1ksQHc' # YouTube 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream """ import argparse import os import sys from pathlib import Path import cv2 import torch import torch.backends.cudnn as cudnn FILE = Path(__file__).resolve() ROOT = FILE.parents[0] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.common import DetectMultiBackend from utils.datasets import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams from utils.general import (LOGGER, check_file, check_img_size, check_imshow, check_requirements, colorstr, increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box from utils.torch_utils import select_device, time_sync @torch.no_grad() def run(weights=ROOT / 'yolov5s.pt', # model.pt path(s) source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam imgsz=(640, 640), # inference size (height, width) conf_thres=0.25, # confidence threshold iou_thres=0.45, # NMS IOU threshold max_det=1000, # maximum detections per image device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu view_img=False, # show results save_txt=False, # save results to *.txt save_conf=False, # save confidences in --save-txt labels save_crop=False, # save cropped prediction boxes nosave=False, # do not save images/videos classes=None, # filter by class: --class 0, or --class 0 2 3 agnostic_nms=False, # class-agnostic NMS augment=False, # augmented inference visualize=False, # visualize features update=False, # update all models project=ROOT / 'runs/detect', # save results to project/name name='exp', # save results to project/name exist_ok=False, # existing project/name ok, do not increment line_thickness=3, # bounding box thickness (pixels) hide_labels=False, # hide labels hide_conf=False, # hide confidences half=False, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference ): source = str(source) save_img = not nosave and not source.endswith('.txt') # save inference images is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://')) webcam = source.isnumeric() or source.endswith('.txt') or (is_url and not is_file) if is_url and is_file: source = check_file(source) # download # Directories save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir # Load model device = select_device(device) model = DetectMultiBackend(weights, device=device, dnn=dnn) stride, names, pt, jit, onnx, engine = model.stride, model.names, model.pt, model.jit, model.onnx, model.engine imgsz = check_img_size(imgsz, s=stride) # check image size # Half half &= (pt or jit or engine) and device.type != 'cpu' # half precision only supported by PyTorch on CUDA if pt or jit: model.model.half() if half else model.model.float() # Dataloader if webcam: view_img = check_imshow() cudnn.benchmark = True # set True to speed up constant image size inference dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt) bs = len(dataset) # batch_size else: dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt) bs = 1 # batch_size vid_path, vid_writer = [None] * bs, [None] * bs # Run inference model.warmup(imgsz=(1, 3, *imgsz), half=half) # warmup dt, seen = [0.0, 0.0, 0.0], 0 for path, im, im0s, vid_cap, s in dataset: t1 = time_sync() im = torch.from_numpy(im).to(device) im = im.half() if half else im.float() # uint8 to fp16/32 im /= 255 # 0 - 255 to 0.0 - 1.0 if len(im.shape) == 3: im = im[None] # expand for batch dim t2 = time_sync() dt[0] += t2 - t1 # Inference visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if visualize else False pred = model(im, augment=augment, visualize=visualize) t3 = time_sync() dt[1] += t3 - t2 # NMS pred = non_max_suppression(pred, conf_thres, iou_thres, classes, agnostic_nms, max_det=max_det) dt[2] += time_sync() - t3 # Second-stage classifier (optional) # pred = utils.general.apply_classifier(pred, classifier_model, im, im0s) # Process predictions for i, det in enumerate(pred): # per image seen += 1 if webcam: # batch_size >= 1 p, im0, frame = path[i], im0s[i].copy(), dataset.count s += f'{i}: ' else: p, im0, frame = path, im0s.copy(), getattr(dataset, 'frame', 0) p = Path(p) # to Path save_path = str(save_dir / p.name) # im.jpg txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}') # im.txt s += '%gx%g ' % im.shape[2:] # print string gn = torch.tensor(im0.shape)[[1, 0, 1, 0]] # normalization gain whwh imc = im0.copy() if save_crop else im0 # for save_crop annotator = Annotator(im0, line_width=line_thickness, example=str(names)) if len(det): # Rescale boxes from img_size to im0 size det[:, :4] = scale_coords(im.shape[2:], det[:, :4], im0.shape).round() # Print results for c in det[:, -1].unique(): n = (det[:, -1] == c).sum() # detections per class s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string # Write results for *xyxy, conf, cls in reversed(det): if save_txt: # Write to file xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh line = (cls, *xywh, conf) if save_conf else (cls, *xywh) # label format with open(txt_path + '.txt', 'a') as f: f.write(('%g ' * len(line)).rstrip() % line + '\n') if save_img or save_crop or view_img: # Add bbox to image c = int(cls) # integer class label = None if hide_labels else (names[c] if hide_conf else f'{names[c]} {conf:.2f}') annotator.box_label(xyxy, label, color=colors(c, True)) if save_crop: save_one_box(xyxy, imc, file=save_dir / 'crops' / names[c] / f'{p.stem}.jpg', BGR=True) # Print time (inference-only) LOGGER.info(f'{s}Done. ({t3 - t2:.3f}s)') # Stream results im0 = annotator.result() if view_img: cv2.imshow(str(p), im0) cv2.waitKey(1) # 1 millisecond # Save results (image with detections) if save_img: if dataset.mode == 'image': cv2.imwrite(save_path, im0) else: # 'video' or 'stream' if vid_path[i] != save_path: # new video vid_path[i] = save_path if isinstance(vid_writer[i], cv2.VideoWriter): vid_writer[i].release() # release previous video writer if vid_cap: # video fps = vid_cap.get(cv2.CAP_PROP_FPS) w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH)) h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) else: # stream fps, w, h = 30, im0.shape[1], im0.shape[0] save_path += '.mp4' vid_writer[i] = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h)) vid_writer[i].write(im0) # Print results t = tuple(x / seen * 1E3 for x in dt) # speeds per image LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {(1, 3, *imgsz)}' % t) if save_txt or save_img: s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else '' LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}{s}") if update: strip_optimizer(weights) # update model (to fix SourceChangeWarning) def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model path(s)') parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob, 0 for webcam') parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w') parser.add_argument('--conf-thres', type=float, default=0.25, help='confidence threshold') parser.add_argument('--iou-thres', type=float, default=0.45, help='NMS IoU threshold') parser.add_argument('--max-det', type=int, default=1000, help='maximum detections per image') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--view-img', action='store_true', help='show results') parser.add_argument('--save-txt', action='store_true', help='save results to *.txt') parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels') parser.add_argument('--save-crop', action='store_true', help='save cropped prediction boxes') parser.add_argument('--nosave', action='store_true', help='do not save images/videos') parser.add_argument('--classes', nargs='+', type=int, help='filter by class: --classes 0, or --classes 0 2 3') parser.add_argument('--agnostic-nms', action='store_true', help='class-agnostic NMS') parser.add_argument('--augment', action='store_true', help='augmented inference') parser.add_argument('--visualize', action='store_true', help='visualize features') parser.add_argument('--update', action='store_true', help='update all models') parser.add_argument('--project', default=ROOT / 'runs/detect', help='save results to project/name') parser.add_argument('--name', default='exp', help='save results to project/name') parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') parser.add_argument('--line-thickness', default=3, type=int, help='bounding box thickness (pixels)') parser.add_argument('--hide-labels', default=False, action='store_true', help='hide labels') parser.add_argument('--hide-conf', default=False, action='store_true', help='hide confidences') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') opt = parser.parse_args() opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand print_args(FILE.stem, opt) return opt def main(opt): check_requirements(exclude=('tensorboard', 'thop')) run(**vars(opt)) if __name__ == "__main__": opt = parse_opt() main(opt) ================================================ FILE: export.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Export a YOLOv5 PyTorch model to other formats. TensorFlow exports authored by https://github.com/zldrobit Format | Example | `--include ...` argument --- | --- | --- PyTorch | yolov5s.pt | - TorchScript | yolov5s.torchscript | `torchscript` ONNX | yolov5s.onnx | `onnx` CoreML | yolov5s.mlmodel | `coreml` TensorFlow SavedModel | yolov5s_saved_model/ | `saved_model` TensorFlow GraphDef | yolov5s.pb | `pb` TensorFlow Lite | yolov5s.tflite | `tflite` TensorFlow.js | yolov5s_web_model/ | `tfjs` TensorRT | yolov5s.engine | `engine` Usage: $ python path/to/export.py --weights yolov5s.pt --include torchscript onnx coreml saved_model pb tflite tfjs Inference: $ python path/to/detect.py --weights yolov5s.pt yolov5s.torchscript yolov5s.onnx yolov5s.mlmodel (under development) yolov5s_saved_model yolov5s.pb yolov5s.tflite yolov5s.engine TensorFlow.js: $ cd .. && git clone https://github.com/zldrobit/tfjs-yolov5-example.git && cd tfjs-yolov5-example $ npm install $ ln -s ../../yolov5/yolov5s_web_model public/yolov5s_web_model $ npm start """ import argparse import json import os import subprocess import sys import time from pathlib import Path # activate rknn hack if len(sys.argv)>=3 and '--rknpu' in sys.argv: _index = sys.argv.index('--rknpu') if sys.argv[_index+1].upper() in ['RK1808', 'RV1109', 'RV1126','RK3399PRO']: os.environ['RKNN_model_hack'] = 'npu_1' elif sys.argv[_index+1].upper() in ['RK3566', 'RK3568', 'RK3588','RK3588S']: os.environ['RKNN_model_hack'] = 'npu_2' else: assert False,"{} not recognized".format(sys.argv[_index+1]) import torch import torch.nn as nn from torch.utils.mobile_optimizer import optimize_for_mobile FILE = Path(__file__).resolve() ROOT = FILE.parents[0] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.common import Conv from models.experimental import attempt_load from models.yolo import Detect from utils.activations import SiLU from utils.datasets import LoadImages from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, colorstr, file_size, print_args, url2file) from utils.torch_utils import select_device def export_torchscript(model, im, file, optimize, prefix=colorstr('TorchScript:')): # YOLOv5 TorchScript model export try: LOGGER.info(f'\n{prefix} starting export with torch {torch.__version__}...') f = file.with_suffix('.torchscript') ts = torch.jit.trace(model, im, strict=False) d = {"shape": im.shape, "stride": int(max(model.stride)), "names": model.names} extra_files = {'config.txt': json.dumps(d)} # torch._C.ExtraFilesMap() (optimize_for_mobile(ts) if optimize else ts).save(str(f), _extra_files=extra_files) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') except Exception as e: LOGGER.info(f'{prefix} export failure: {e}') def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorstr('ONNX:')): # YOLOv5 ONNX export try: check_requirements(('onnx',)) import onnx LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...') f = file.with_suffix('.onnx') torch.onnx.export(model, im, f, verbose=False, opset_version=opset, training=torch.onnx.TrainingMode.TRAINING if train else torch.onnx.TrainingMode.EVAL, do_constant_folding=not train, input_names=['images'], output_names=['output'], dynamic_axes={'images': {0: 'batch', 2: 'height', 3: 'width'}, # shape(1,3,640,640) 'output': {0: 'batch', 1: 'anchors'} # shape(1,25200,85) } if dynamic else None) # Checks model_onnx = onnx.load(f) # load onnx model onnx.checker.check_model(model_onnx) # check onnx model # LOGGER.info(onnx.helper.printable_graph(model_onnx.graph)) # print # Simplify if simplify: try: check_requirements(('onnx-simplifier',)) import onnxsim LOGGER.info(f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...') model_onnx, check = onnxsim.simplify( model_onnx, dynamic_input_shape=dynamic, input_shapes={'images': list(im.shape)} if dynamic else None) assert check, 'assert check failed' onnx.save(model_onnx, f) except Exception as e: LOGGER.info(f'{prefix} simplifier failure: {e}') LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') LOGGER.info(f"{prefix} run --dynamic ONNX model inference with: 'python detect.py --weights {f}'") except Exception as e: LOGGER.info(f'{prefix} export failure: {e}') def export_coreml(model, im, file, prefix=colorstr('CoreML:')): # YOLOv5 CoreML export ct_model = None try: check_requirements(('coremltools',)) import coremltools as ct LOGGER.info(f'\n{prefix} starting export with coremltools {ct.__version__}...') f = file.with_suffix('.mlmodel') model.train() # CoreML exports should be placed in model.train() mode ts = torch.jit.trace(model, im, strict=False) # TorchScript model ct_model = ct.convert(ts, inputs=[ct.ImageType('image', shape=im.shape, scale=1 / 255, bias=[0, 0, 0])]) ct_model.save(f) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') except Exception as e: LOGGER.info(f'\n{prefix} export failure: {e}') return ct_model def export_saved_model(model, im, file, dynamic, tf_nms=False, agnostic_nms=False, topk_per_class=100, topk_all=100, iou_thres=0.45, conf_thres=0.25, prefix=colorstr('TensorFlow saved_model:')): # YOLOv5 TensorFlow saved_model export keras_model = None try: import tensorflow as tf from tensorflow import keras from models.tf import TFDetect, TFModel LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') f = str(file).replace('.pt', '_saved_model') batch_size, ch, *imgsz = list(im.shape) # BCHW tf_model = TFModel(cfg=model.yaml, model=model, nc=model.nc, imgsz=imgsz) im = tf.zeros((batch_size, *imgsz, 3)) # BHWC order for TensorFlow y = tf_model.predict(im, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres) inputs = keras.Input(shape=(*imgsz, 3), batch_size=None if dynamic else batch_size) outputs = tf_model.predict(inputs, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres) keras_model = keras.Model(inputs=inputs, outputs=outputs) keras_model.trainable = False keras_model.summary() keras_model.save(f, save_format='tf') LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') except Exception as e: LOGGER.info(f'\n{prefix} export failure: {e}') return keras_model def export_pb(keras_model, im, file, prefix=colorstr('TensorFlow GraphDef:')): # YOLOv5 TensorFlow GraphDef *.pb export https://github.com/leimao/Frozen_Graph_TensorFlow try: import tensorflow as tf from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2 LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') f = file.with_suffix('.pb') m = tf.function(lambda x: keras_model(x)) # full model m = m.get_concrete_function(tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype)) frozen_func = convert_variables_to_constants_v2(m) frozen_func.graph.as_graph_def() tf.io.write_graph(graph_or_graph_def=frozen_func.graph, logdir=str(f.parent), name=f.name, as_text=False) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') except Exception as e: LOGGER.info(f'\n{prefix} export failure: {e}') def export_tflite(keras_model, im, file, int8, data, ncalib, prefix=colorstr('TensorFlow Lite:')): # YOLOv5 TensorFlow Lite export try: import tensorflow as tf from models.tf import representative_dataset_gen LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') batch_size, ch, *imgsz = list(im.shape) # BCHW f = str(file).replace('.pt', '-fp16.tflite') converter = tf.lite.TFLiteConverter.from_keras_model(keras_model) converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS] converter.target_spec.supported_types = [tf.float16] converter.optimizations = [tf.lite.Optimize.DEFAULT] if int8: dataset = LoadImages(check_dataset(data)['train'], img_size=imgsz, auto=False) # representative data converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib) converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] converter.target_spec.supported_types = [] converter.inference_input_type = tf.uint8 # or tf.int8 converter.inference_output_type = tf.uint8 # or tf.int8 converter.experimental_new_quantizer = False f = str(file).replace('.pt', '-int8.tflite') tflite_model = converter.convert() open(f, "wb").write(tflite_model) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') except Exception as e: LOGGER.info(f'\n{prefix} export failure: {e}') def export_tfjs(keras_model, im, file, prefix=colorstr('TensorFlow.js:')): # YOLOv5 TensorFlow.js export try: check_requirements(('tensorflowjs',)) import re import tensorflowjs as tfjs LOGGER.info(f'\n{prefix} starting export with tensorflowjs {tfjs.__version__}...') f = str(file).replace('.pt', '_web_model') # js dir f_pb = file.with_suffix('.pb') # *.pb path f_json = f + '/model.json' # *.json path cmd = f"tensorflowjs_converter --input_format=tf_frozen_model " \ f"--output_node_names='Identity,Identity_1,Identity_2,Identity_3' {f_pb} {f}" subprocess.run(cmd, shell=True) json = open(f_json).read() with open(f_json, 'w') as j: # sort JSON Identity_* in ascending order subst = re.sub( r'{"outputs": {"Identity.?.?": {"name": "Identity.?.?"}, ' r'"Identity.?.?": {"name": "Identity.?.?"}, ' r'"Identity.?.?": {"name": "Identity.?.?"}, ' r'"Identity.?.?": {"name": "Identity.?.?"}}}', r'{"outputs": {"Identity": {"name": "Identity"}, ' r'"Identity_1": {"name": "Identity_1"}, ' r'"Identity_2": {"name": "Identity_2"}, ' r'"Identity_3": {"name": "Identity_3"}}}', json) j.write(subst) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') except Exception as e: LOGGER.info(f'\n{prefix} export failure: {e}') def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=False, prefix=colorstr('TensorRT:')): try: check_requirements(('tensorrt',)) import tensorrt as trt opset = (12, 13)[trt.__version__[0] == '8'] # test on TensorRT 7.x and 8.x export_onnx(model, im, file, opset, train, False, simplify) onnx = file.with_suffix('.onnx') assert onnx.exists(), f'failed to export ONNX file: {onnx}' LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...') f = file.with_suffix('.engine') # TensorRT engine file logger = trt.Logger(trt.Logger.INFO) if verbose: logger.min_severity = trt.Logger.Severity.VERBOSE builder = trt.Builder(logger) config = builder.create_builder_config() config.max_workspace_size = workspace * 1 << 30 flag = (1 << int(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH)) network = builder.create_network(flag) parser = trt.OnnxParser(network, logger) if not parser.parse_from_file(str(onnx)): raise RuntimeError(f'failed to load ONNX file: {onnx}') inputs = [network.get_input(i) for i in range(network.num_inputs)] outputs = [network.get_output(i) for i in range(network.num_outputs)] LOGGER.info(f'{prefix} Network Description:') for inp in inputs: LOGGER.info(f'{prefix}\tinput "{inp.name}" with shape {inp.shape} and dtype {inp.dtype}') for out in outputs: LOGGER.info(f'{prefix}\toutput "{out.name}" with shape {out.shape} and dtype {out.dtype}') half &= builder.platform_has_fast_fp16 LOGGER.info(f'{prefix} building FP{16 if half else 32} engine in {f}') if half: config.set_flag(trt.BuilderFlag.FP16) with builder.build_engine(network, config) as engine, open(f, 'wb') as t: t.write(engine.serialize()) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') except Exception as e: LOGGER.info(f'\n{prefix} export failure: {e}') @torch.no_grad() def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' weights=ROOT / 'yolov5s.pt', # weights path imgsz=(640, 640), # image (height, width) batch_size=1, # batch size device='cpu', # cuda device, i.e. 0 or 0,1,2,3 or cpu include=('torchscript', 'onnx', 'coreml'), # include formats half=False, # FP16 half-precision export inplace=False, # set YOLOv5 Detect() inplace=True train=False, # model.train() mode optimize=False, # TorchScript: optimize for mobile int8=False, # CoreML/TF INT8 quantization dynamic=False, # ONNX/TF: dynamic axes simplify=False, # ONNX: simplify model opset=14, # ONNX: opset version verbose=False, # TensorRT: verbose log workspace=4, # TensorRT: workspace size (GB) nms=False, # TF: add NMS to model agnostic_nms=False, # TF: add agnostic NMS to model topk_per_class=100, # TF.js NMS: topk per class to keep topk_all=100, # TF.js NMS: topk for all classes to keep iou_thres=0.45, # TF.js NMS: IoU threshold conf_thres=0.25, # TF.js NMS: confidence threshold rknn_friendly = False, ): t = time.time() include = [x.lower() for x in include] tf_exports = list(x in include for x in ('saved_model', 'pb', 'tflite', 'tfjs')) # TensorFlow exports imgsz *= 2 if len(imgsz) == 1 else 1 # expand file = Path(url2file(weights) if str(weights).startswith(('http:/', 'https:/')) else weights) # Load PyTorch model device = select_device(device) assert not (device.type == 'cpu' and half), '--half only compatible with GPU export, i.e. use --device 0' model = attempt_load(weights, map_location=device, inplace=True, fuse=True) # load FP32 model nc, names = model.nc, model.names # number of classes, class names # Input gs = int(max(model.stride)) # grid size (max stride) imgsz = [check_img_size(x, gs) for x in imgsz] # verify img_size are gs-multiples im = torch.zeros(batch_size, 3, *imgsz).to(device) # image size(1,3,320,192) BCHW iDetection # Update model if half: im, model = im.half(), model.half() # to FP16 model.train() if train else model.eval() # training mode = no Detect() layer grid construction for k, m in model.named_modules(): if isinstance(m, Conv): # assign export-friendly activations if isinstance(m.act, nn.SiLU): m.act = SiLU() elif isinstance(m, Detect): m.inplace = inplace m.onnx_dynamic = dynamic # m.forward = m.forward_export # assign forward (optional) if os.getenv('RKNN_model_hack', '0') == 'npu_1': from models.common import Focus from models.common_rk_plug_in import surrogate_focus if isinstance(model.model[0], Focus): # For yolo v5 version surrogate_focous = surrogate_focus(int(model.model[0].conv.conv.weight.shape[1]/4), model.model[0].conv.conv.weight.shape[0], k=tuple(model.model[0].conv.conv.weight.shape[2:4]), s=model.model[0].conv.conv.stride, p=model.model[0].conv.conv.padding, g=model.model[0].conv.conv.groups, act=True) surrogate_focous.conv.conv.weight = model.model[0].conv.conv.weight surrogate_focous.conv.conv.bias = model.model[0].conv.conv.bias surrogate_focous.conv.act = model.model[0].conv.act temp_i = model.model[0].i temp_f = model.model[0].f model.model[0] = surrogate_focous model.model[0].i = temp_i model.model[0].f = temp_f model.model[0].eval() elif isinstance(model.model[0], Conv) and model.model[0].conv.kernel_size == (6, 6): # For yolo v6 version surrogate_focous = surrogate_focus(model.model[0].conv.weight.shape[1], model.model[0].conv.weight.shape[0], k=(3,3), # 6/2, 6/2 s=1, p=(1,1), # 2/2, 2/2 g=model.model[0].conv.groups, act=hasattr(model.model[0], 'act')) surrogate_focous.conv.conv.weight[:,:3,:,:] = model.model[0].conv.weight[:,:,::2,::2] surrogate_focous.conv.conv.weight[:,3:6,:,:] = model.model[0].conv.weight[:,:,1::2,::2] surrogate_focous.conv.conv.weight[:,6:9,:,:] = model.model[0].conv.weight[:,:,::2,1::2] surrogate_focous.conv.conv.weight[:,9:,:,:] = model.model[0].conv.weight[:,:,1::2,1::2] surrogate_focous.conv.conv.bias = model.model[0].conv.bias surrogate_focous.conv.act = model.model[0].act temp_i = model.model[0].i temp_f = model.model[0].f model.model[0] = surrogate_focous model.model[0].i = temp_i model.model[0].f = temp_f model.model[0].eval() # save anchors if isinstance(model.model[-1], Detect): print('---> save anchors for RKNN') RK_anchors = model.model[-1].stride.reshape(3,1).repeat(1,3).reshape(-1,1)* model.model[-1].anchors.reshape(9,2) RK_anchors = RK_anchors.tolist() print(RK_anchors) with open(file.with_suffix('.anchors.txt'), 'w') as anf: anf.write(str(RK_anchors)) for _ in range(2): y = model(im) # dry runs LOGGER.info(f"\n{colorstr('PyTorch:')} starting from {file} ({file_size(file):.1f} MB)") # Exports if 'torchscript' in include: export_torchscript(model, im, file, optimize) if 'onnx' in include: export_onnx(model, im, file, opset, train, dynamic, simplify) if 'engine' in include: export_engine(model, im, file, train, half, simplify, workspace, verbose) if 'coreml' in include: export_coreml(model, im, file) # TensorFlow Exports if any(tf_exports): pb, tflite, tfjs = tf_exports[1:] assert not (tflite and tfjs), 'TFLite and TF.js models must be exported separately, please pass only one type.' model = export_saved_model(model, im, file, dynamic, tf_nms=nms or agnostic_nms or tfjs, agnostic_nms=agnostic_nms or tfjs, topk_per_class=topk_per_class, topk_all=topk_all, conf_thres=conf_thres, iou_thres=iou_thres) # keras model if pb or tfjs: # pb prerequisite to tfjs export_pb(model, im, file) if tflite: export_tflite(model, im, file, int8=int8, data=data, ncalib=100) if tfjs: export_tfjs(model, im, file) # Finish LOGGER.info(f'\nExport complete ({time.time() - t:.2f}s)' f"\nResults saved to {colorstr('bold', file.parent.resolve())}" f'\nVisualize with https://netron.app') def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model.pt path(s)') parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640, 640], help='image (h, w)') parser.add_argument('--batch-size', type=int, default=1, help='batch size') parser.add_argument('--device', default='cpu', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--half', action='store_true', help='FP16 half-precision export') parser.add_argument('--inplace', action='store_true', help='set YOLOv5 Detect() inplace=True') parser.add_argument('--train', action='store_true', help='model.train() mode') parser.add_argument('--optimize', action='store_true', help='TorchScript: optimize for mobile') parser.add_argument('--int8', action='store_true', help='CoreML/TF INT8 quantization') parser.add_argument('--dynamic', action='store_true', help='ONNX/TF: dynamic axes') parser.add_argument('--simplify', action='store_true', help='ONNX: simplify model') parser.add_argument('--opset', type=int, default=12, help='ONNX: opset version') parser.add_argument('--verbose', action='store_true', help='TensorRT: verbose log') parser.add_argument('--workspace', type=int, default=4, help='TensorRT: workspace size (GB)') parser.add_argument('--nms', action='store_true', help='TF: add NMS to model') parser.add_argument('--agnostic-nms', action='store_true', help='TF: add agnostic NMS to model') parser.add_argument('--topk-per-class', type=int, default=100, help='TF.js NMS: topk per class to keep') parser.add_argument('--topk-all', type=int, default=100, help='TF.js NMS: topk for all classes to keep') parser.add_argument('--iou-thres', type=float, default=0.45, help='TF.js NMS: IoU threshold') parser.add_argument('--conf-thres', type=float, default=0.25, help='TF.js NMS: confidence threshold') parser.add_argument('--include', nargs='+', default=['torchscript'], help='available formats are (torchscript, onnx, engine, coreml, saved_model, pb, tflite, tfjs)') parser.add_argument('--rknpu', default=None, help='RKNN npu platform') opt = parser.parse_args() print_args(FILE.stem, opt) return opt def main(opt): for opt.weights in (opt.weights if isinstance(opt.weights, list) else [opt.weights]): run(**vars(opt)) if __name__ == "__main__": opt = parse_opt() del opt.rknpu main(opt) ================================================ FILE: hubconf.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ PyTorch Hub models https://pytorch.org/hub/ultralytics_yolov5/ Usage: import torch model = torch.hub.load('ultralytics/yolov5', 'yolov5s') model = torch.hub.load('ultralytics/yolov5:master', 'custom', 'path/to/yolov5s.onnx') # file from branch """ import torch def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): """Creates a specified YOLOv5 model Arguments: name (str): name of model, i.e. 'yolov5s' pretrained (bool): load pretrained weights into the model channels (int): number of input channels classes (int): number of model classes autoshape (bool): apply YOLOv5 .autoshape() wrapper to model verbose (bool): print all information to screen device (str, torch.device, None): device to use for model parameters Returns: YOLOv5 pytorch model """ from pathlib import Path from models.common import AutoShape, DetectMultiBackend from models.yolo import Model from utils.downloads import attempt_download from utils.general import check_requirements, intersect_dicts, set_logging from utils.torch_utils import select_device check_requirements(exclude=('tensorboard', 'thop', 'opencv-python')) set_logging(verbose=verbose) name = Path(name) path = name.with_suffix('.pt') if name.suffix == '' else name # checkpoint path try: device = select_device(('0' if torch.cuda.is_available() else 'cpu') if device is None else device) if pretrained and channels == 3 and classes == 80: model = DetectMultiBackend(path, device=device) # download/load FP32 model # model = models.experimental.attempt_load(path, map_location=device) # download/load FP32 model else: cfg = list((Path(__file__).parent / 'models').rglob(f'{path.stem}.yaml'))[0] # model.yaml path model = Model(cfg, channels, classes) # create model if pretrained: ckpt = torch.load(attempt_download(path), map_location=device) # load csd = ckpt['model'].float().state_dict() # checkpoint state_dict as FP32 csd = intersect_dicts(csd, model.state_dict(), exclude=['anchors']) # intersect model.load_state_dict(csd, strict=False) # load if len(ckpt['model'].names) == classes: model.names = ckpt['model'].names # set class names attribute if autoshape: model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS return model.to(device) except Exception as e: help_url = 'https://github.com/ultralytics/yolov5/issues/36' s = 'Cache may be out of date, try `force_reload=True`. See %s for help.' % help_url raise Exception(s) from e def custom(path='path/to/model.pt', autoshape=True, verbose=True, device=None): # YOLOv5 custom or local model return _create(path, autoshape=autoshape, verbose=verbose, device=device) def yolov5n(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-nano model https://github.com/ultralytics/yolov5 return _create('yolov5n', pretrained, channels, classes, autoshape, verbose, device) def yolov5s(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-small model https://github.com/ultralytics/yolov5 return _create('yolov5s', pretrained, channels, classes, autoshape, verbose, device) def yolov5m(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-medium model https://github.com/ultralytics/yolov5 return _create('yolov5m', pretrained, channels, classes, autoshape, verbose, device) def yolov5l(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-large model https://github.com/ultralytics/yolov5 return _create('yolov5l', pretrained, channels, classes, autoshape, verbose, device) def yolov5x(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-xlarge model https://github.com/ultralytics/yolov5 return _create('yolov5x', pretrained, channels, classes, autoshape, verbose, device) def yolov5n6(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-nano-P6 model https://github.com/ultralytics/yolov5 return _create('yolov5n6', pretrained, channels, classes, autoshape, verbose, device) def yolov5s6(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-small-P6 model https://github.com/ultralytics/yolov5 return _create('yolov5s6', pretrained, channels, classes, autoshape, verbose, device) def yolov5m6(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-medium-P6 model https://github.com/ultralytics/yolov5 return _create('yolov5m6', pretrained, channels, classes, autoshape, verbose, device) def yolov5l6(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-large-P6 model https://github.com/ultralytics/yolov5 return _create('yolov5l6', pretrained, channels, classes, autoshape, verbose, device) def yolov5x6(pretrained=True, channels=3, classes=80, autoshape=True, verbose=True, device=None): # YOLOv5-xlarge-P6 model https://github.com/ultralytics/yolov5 return _create('yolov5x6', pretrained, channels, classes, autoshape, verbose, device) if __name__ == '__main__': model = _create(name='yolov5s', pretrained=True, channels=3, classes=80, autoshape=True, verbose=True) # pretrained # model = custom(path='path/to/model.pt') # custom # Verify inference from pathlib import Path import cv2 import numpy as np from PIL import Image imgs = ['data/images/zidane.jpg', # filename Path('data/images/zidane.jpg'), # Path 'https://ultralytics.com/images/zidane.jpg', # URI cv2.imread('data/images/bus.jpg')[:, :, ::-1], # OpenCV Image.open('data/images/bus.jpg'), # PIL np.zeros((320, 640, 3))] # numpy results = model(imgs, size=320) # batched inference results.print() results.save() ================================================ FILE: models/__init__.py ================================================ ================================================ FILE: models/common.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Common modules """ import os import json import math import platform import warnings from collections import OrderedDict, namedtuple from copy import copy from pathlib import Path import cv2 import numpy as np import pandas as pd import requests import torch import torch.nn as nn from PIL import Image from torch.cuda import amp from utils.datasets import exif_transpose, letterbox from utils.general import (LOGGER, check_requirements, check_suffix, colorstr, increment_path, make_divisible, non_max_suppression, scale_coords, xywh2xyxy, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box from utils.torch_utils import copy_attr, time_sync def autopad(k, p=None): # kernel, padding # Pad to 'same' if p is None: p = k // 2 if isinstance(k, int) else [x // 2 for x in k] # auto-pad return p class Conv(nn.Module): # Standard convolution def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups super().__init__() self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g, bias=False) self.bn = nn.BatchNorm2d(c2) self.act = nn.ReLU() if act is True else (act if isinstance(act, nn.Module) else nn.Identity()) def forward(self, x): return self.act(self.bn(self.conv(x))) def forward_fuse(self, x): return self.act(self.conv(x)) class DWConv(Conv): # Depth-wise convolution class def __init__(self, c1, c2, k=1, s=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups super().__init__(c1, c2, k, s, g=math.gcd(c1, c2), act=act) class TransformerLayer(nn.Module): # Transformer layer https://arxiv.org/abs/2010.11929 (LayerNorm layers removed for better performance) def __init__(self, c, num_heads): super().__init__() self.q = nn.Linear(c, c, bias=False) self.k = nn.Linear(c, c, bias=False) self.v = nn.Linear(c, c, bias=False) self.ma = nn.MultiheadAttention(embed_dim=c, num_heads=num_heads) self.fc1 = nn.Linear(c, c, bias=False) self.fc2 = nn.Linear(c, c, bias=False) def forward(self, x): x = self.ma(self.q(x), self.k(x), self.v(x))[0] + x x = self.fc2(self.fc1(x)) + x return x class TransformerBlock(nn.Module): # Vision Transformer https://arxiv.org/abs/2010.11929 def __init__(self, c1, c2, num_heads, num_layers): super().__init__() self.conv = None if c1 != c2: self.conv = Conv(c1, c2) self.linear = nn.Linear(c2, c2) # learnable position embedding self.tr = nn.Sequential(*(TransformerLayer(c2, num_heads) for _ in range(num_layers))) self.c2 = c2 def forward(self, x): if self.conv is not None: x = self.conv(x) b, _, w, h = x.shape p = x.flatten(2).permute(2, 0, 1) return self.tr(p + self.linear(p)).permute(1, 2, 0).reshape(b, self.c2, w, h) class Bottleneck(nn.Module): # Standard bottleneck def __init__(self, c1, c2, shortcut=True, g=1, e=0.5): # ch_in, ch_out, shortcut, groups, expansion super().__init__() c_ = int(c2 * e) # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c_, c2, 3, 1, g=g) self.add = shortcut and c1 == c2 def forward(self, x): return x + self.cv2(self.cv1(x)) if self.add else self.cv2(self.cv1(x)) class BottleneckCSP(nn.Module): # CSP Bottleneck https://github.com/WongKinYiu/CrossStagePartialNetworks def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, number, shortcut, groups, expansion super().__init__() c_ = int(c2 * e) # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = nn.Conv2d(c1, c_, 1, 1, bias=False) self.cv3 = nn.Conv2d(c_, c_, 1, 1, bias=False) self.cv4 = Conv(2 * c_, c2, 1, 1) self.bn = nn.BatchNorm2d(2 * c_) # applied to cat(cv2, cv3) self.act = nn.ReLU() self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) def forward(self, x): y1 = self.cv3(self.m(self.cv1(x))) y2 = self.cv2(x) return self.cv4(self.act(self.bn(torch.cat((y1, y2), dim=1)))) class C3(nn.Module): # CSP Bottleneck with 3 convolutions def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, number, shortcut, groups, expansion super().__init__() c_ = int(c2 * e) # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c1, c_, 1, 1) self.cv3 = Conv(2 * c_, c2, 1) # act=FReLU(c2) self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) # self.m = nn.Sequential(*[CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n)]) def forward(self, x): return self.cv3(torch.cat((self.m(self.cv1(x)), self.cv2(x)), dim=1)) class C3TR(C3): # C3 module with TransformerBlock() def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): super().__init__(c1, c2, n, shortcut, g, e) c_ = int(c2 * e) self.m = TransformerBlock(c_, c_, 4, n) class C3SPP(C3): # C3 module with SPP() def __init__(self, c1, c2, k=(5, 9, 13), n=1, shortcut=True, g=1, e=0.5): super().__init__(c1, c2, n, shortcut, g, e) c_ = int(c2 * e) self.m = SPP(c_, c_, k) class C3Ghost(C3): # C3 module with GhostBottleneck() def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): super().__init__(c1, c2, n, shortcut, g, e) c_ = int(c2 * e) # hidden channels self.m = nn.Sequential(*(GhostBottleneck(c_, c_) for _ in range(n))) if os.getenv('RKNN_model_hack', '0') == '0': class SPP(nn.Module): # Spatial Pyramid Pooling (SPP) layer https://arxiv.org/abs/1406.4729 def __init__(self, c1, c2, k=(5, 9, 13)): super().__init__() c_ = c1 // 2 # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c_ * (len(k) + 1), c2, 1, 1) self.m = nn.ModuleList([nn.MaxPool2d(kernel_size=x, stride=1, padding=x // 2) for x in k]) def forward(self, x): x = self.cv1(x) with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning return self.cv2(torch.cat([x] + [m(x) for m in self.m], 1)) elif os.getenv('RKNN_model_hack', '0') in ['npu_1', 'npu_2']: # TODO remove this hack when rknn-toolkit1/2 add this optimize rules class SPP(nn.Module): def __init__(self, c1, c2, k=(5, 9, 13)): super().__init__() c_ = c1 // 2 # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c_ * (len(k) + 1), c2, 1, 1) self.m = nn.ModuleList([nn.MaxPool2d(kernel_size=x, stride=1, padding=x // 2) for x in k]) for value in k: assert (value%2 == 1) and (value!= 1), "value in [{}] only support odd number for RKNN model hack" def forward(self, x): x = self.cv1(x) with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning y = [x] for maxpool in self.m: kernel_size = maxpool.kernel_size m = x for i in range(math.floor(kernel_size/2)): m = torch.nn.functional.max_pool2d(m, 3, 1, 1) y = [*y, m] return self.cv2(torch.cat(y, 1)) if os.getenv('RKNN_model_hack', '0') in ['0','npu_2']: class SPPF(nn.Module): # Spatial Pyramid Pooling - Fast (SPPF) layer for YOLOv5 by Glenn Jocher def __init__(self, c1, c2, k=5): # equivalent to SPP(k=(5, 9, 13)) super().__init__() c_ = c1 // 2 # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c_ * 4, c2, 1, 1) self.m = nn.MaxPool2d(kernel_size=k, stride=1, padding=k // 2) def forward(self, x): x = self.cv1(x) with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning y1 = self.m(x) y2 = self.m(y1) return self.cv2(torch.cat([x, y1, y2, self.m(y2)], 1)) elif os.getenv('RKNN_model_hack', '0') == 'npu_1': class SPPF(nn.Module): # Spatial Pyramid Pooling - Fast (SPPF) layer for YOLOv5 by Glenn Jocher def __init__(self, c1, c2, k=5): # equivalent to SPP(k=(5, 9, 13)) super().__init__() c_ = c1 // 2 # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c_ * 4, c2, 1, 1) self.m = nn.MaxPool2d(kernel_size=k, stride=1, padding=k // 2) def forward(self, x): x = self.cv1(x) with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning y1 = self.m(x) y2 = self.m(y1) with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning y = [x] kernel_size = self.m.kernel_size _3x3_stack = math.floor(kernel_size/2) for i in range(3): m = y[-1] for _ in range(_3x3_stack): m = torch.nn.functional.max_pool2d(m, 3, 1, 1) y = [*y, m] return self.cv2(torch.cat(y, 1)) class Focus(nn.Module): # Focus wh information into c-space def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups super().__init__() self.conv = Conv(c1 * 4, c2, k, s, p, g, act) # self.contract = Contract(gain=2) def forward(self, x): # x(b,c,w,h) -> y(b,4c,w/2,h/2) return self.conv(torch.cat([x[..., ::2, ::2], x[..., 1::2, ::2], x[..., ::2, 1::2], x[..., 1::2, 1::2]], 1)) # return self.conv(self.contract(x)) class GhostConv(nn.Module): # Ghost Convolution https://github.com/huawei-noah/ghostnet def __init__(self, c1, c2, k=1, s=1, g=1, act=True): # ch_in, ch_out, kernel, stride, groups super().__init__() c_ = c2 // 2 # hidden channels self.cv1 = Conv(c1, c_, k, s, None, g, act) self.cv2 = Conv(c_, c_, 5, 1, None, c_, act) def forward(self, x): y = self.cv1(x) return torch.cat([y, self.cv2(y)], 1) class GhostBottleneck(nn.Module): # Ghost Bottleneck https://github.com/huawei-noah/ghostnet def __init__(self, c1, c2, k=3, s=1): # ch_in, ch_out, kernel, stride super().__init__() c_ = c2 // 2 self.conv = nn.Sequential(GhostConv(c1, c_, 1, 1), # pw DWConv(c_, c_, k, s, act=False) if s == 2 else nn.Identity(), # dw GhostConv(c_, c2, 1, 1, act=False)) # pw-linear self.shortcut = nn.Sequential(DWConv(c1, c1, k, s, act=False), Conv(c1, c2, 1, 1, act=False)) if s == 2 else nn.Identity() def forward(self, x): return self.conv(x) + self.shortcut(x) class Contract(nn.Module): # Contract width-height into channels, i.e. x(1,64,80,80) to x(1,256,40,40) def __init__(self, gain=2): super().__init__() self.gain = gain def forward(self, x): b, c, h, w = x.size() # assert (h / s == 0) and (W / s == 0), 'Indivisible gain' s = self.gain x = x.view(b, c, h // s, s, w // s, s) # x(1,64,40,2,40,2) x = x.permute(0, 3, 5, 1, 2, 4).contiguous() # x(1,2,2,64,40,40) return x.view(b, c * s * s, h // s, w // s) # x(1,256,40,40) class Expand(nn.Module): # Expand channels into width-height, i.e. x(1,64,80,80) to x(1,16,160,160) def __init__(self, gain=2): super().__init__() self.gain = gain def forward(self, x): b, c, h, w = x.size() # assert C / s ** 2 == 0, 'Indivisible gain' s = self.gain x = x.view(b, s, s, c // s ** 2, h, w) # x(1,2,2,16,80,80) x = x.permute(0, 3, 4, 1, 5, 2).contiguous() # x(1,16,80,2,80,2) return x.view(b, c // s ** 2, h * s, w * s) # x(1,16,160,160) class Concat(nn.Module): # Concatenate a list of tensors along dimension def __init__(self, dimension=1): super().__init__() self.d = dimension def forward(self, x): return torch.cat(x, self.d) class DetectMultiBackend(nn.Module): # YOLOv5 MultiBackend class for python inference on various backends def __init__(self, weights='yolov5s.pt', device=None, dnn=False): # Usage: # PyTorch: weights = *.pt # TorchScript: *.torchscript # CoreML: *.mlmodel # TensorFlow: *_saved_model # TensorFlow: *.pb # TensorFlow Lite: *.tflite # ONNX Runtime: *.onnx # OpenCV DNN: *.onnx with dnn=True # TensorRT: *.engine # RKNN: *.rknn from models.experimental import attempt_download, attempt_load # scoped to avoid circular import super().__init__() w = str(weights[0] if isinstance(weights, list) else weights) suffix = Path(w).suffix.lower() suffixes = ['.pt', '.torchscript', '.onnx', '.engine', '.tflite', '.pb', '', '.mlmodel', '.rknn'] check_suffix(w, suffixes) # check weights have acceptable suffix pt, jit, onnx, engine, tflite, pb, saved_model, coreml, rknn_model = (suffix == x for x in suffixes) # backend booleans stride, names = 64, [f'class{i}' for i in range(1000)] # assign defaults attempt_download(w) # download if not local if jit: # TorchScript LOGGER.info(f'Loading {w} for TorchScript inference...') extra_files = {'config.txt': ''} # model metadata model = torch.jit.load(w, _extra_files=extra_files) if extra_files['config.txt']: d = json.loads(extra_files['config.txt']) # extra_files dict stride, names = int(d['stride']), d['names'] elif pt: # PyTorch model = attempt_load(weights, map_location=device) stride = int(model.stride.max()) # model stride names = model.module.names if hasattr(model, 'module') else model.names # get class names self.model = model # explicitly assign for to(), cpu(), cuda(), half() elif coreml: # CoreML LOGGER.info(f'Loading {w} for CoreML inference...') import coremltools as ct model = ct.models.MLModel(w) elif dnn: # ONNX OpenCV DNN LOGGER.info(f'Loading {w} for ONNX OpenCV DNN inference...') check_requirements(('opencv-python>=4.5.4',)) net = cv2.dnn.readNetFromONNX(w) elif onnx: # ONNX Runtime LOGGER.info(f'Loading {w} for ONNX Runtime inference...') cuda = torch.cuda.is_available() check_requirements(('onnx', 'onnxruntime-gpu' if cuda else 'onnxruntime')) import onnxruntime providers = ['CUDAExecutionProvider', 'CPUExecutionProvider'] if cuda else ['CPUExecutionProvider'] session = onnxruntime.InferenceSession(w, providers=providers) elif engine: # TensorRT LOGGER.info(f'Loading {w} for TensorRT inference...') import tensorrt as trt # https://developer.nvidia.com/nvidia-tensorrt-download Binding = namedtuple('Binding', ('name', 'dtype', 'shape', 'data', 'ptr')) logger = trt.Logger(trt.Logger.INFO) with open(w, 'rb') as f, trt.Runtime(logger) as runtime: model = runtime.deserialize_cuda_engine(f.read()) bindings = OrderedDict() for index in range(model.num_bindings): name = model.get_binding_name(index) dtype = trt.nptype(model.get_binding_dtype(index)) shape = tuple(model.get_binding_shape(index)) data = torch.from_numpy(np.empty(shape, dtype=np.dtype(dtype))).to(device) bindings[name] = Binding(name, dtype, shape, data, int(data.data_ptr())) binding_addrs = OrderedDict((n, d.ptr) for n, d in bindings.items()) context = model.create_execution_context() batch_size = bindings['images'].shape[0] elif rknn_model: # TODO if post-process in model, then we can add code here. pass else: # TensorFlow model (TFLite, pb, saved_model) if pb: # https://www.tensorflow.org/guide/migrate#a_graphpb_or_graphpbtxt LOGGER.info(f'Loading {w} for TensorFlow *.pb inference...') import tensorflow as tf def wrap_frozen_graph(gd, inputs, outputs): x = tf.compat.v1.wrap_function(lambda: tf.compat.v1.import_graph_def(gd, name=""), []) # wrapped return x.prune(tf.nest.map_structure(x.graph.as_graph_element, inputs), tf.nest.map_structure(x.graph.as_graph_element, outputs)) graph_def = tf.Graph().as_graph_def() graph_def.ParseFromString(open(w, 'rb').read()) frozen_func = wrap_frozen_graph(gd=graph_def, inputs="x:0", outputs="Identity:0") elif saved_model: LOGGER.info(f'Loading {w} for TensorFlow saved_model inference...') import tensorflow as tf model = tf.keras.models.load_model(w) elif tflite: # https://www.tensorflow.org/lite/guide/python#install_tensorflow_lite_for_python if 'edgetpu' in w.lower(): LOGGER.info(f'Loading {w} for TensorFlow Lite Edge TPU inference...') import tflite_runtime.interpreter as tfli delegate = {'Linux': 'libedgetpu.so.1', # install https://coral.ai/software/#edgetpu-runtime 'Darwin': 'libedgetpu.1.dylib', 'Windows': 'edgetpu.dll'}[platform.system()] interpreter = tfli.Interpreter(model_path=w, experimental_delegates=[tfli.load_delegate(delegate)]) else: LOGGER.info(f'Loading {w} for TensorFlow Lite inference...') import tensorflow as tf interpreter = tf.lite.Interpreter(model_path=w) # load TFLite model interpreter.allocate_tensors() # allocate input_details = interpreter.get_input_details() # inputs output_details = interpreter.get_output_details() # outputs self.__dict__.update(locals()) # assign all variables to self def forward(self, im, augment=False, visualize=False, val=False): # YOLOv5 MultiBackend inference b, ch, h, w = im.shape # batch, channel, height, width if self.pt: # PyTorch y = self.model(im) if self.jit else self.model(im, augment=augment, visualize=visualize) return y if val else y[0] elif self.coreml: # CoreML im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) im = Image.fromarray((im[0] * 255).astype('uint8')) # im = im.resize((192, 320), Image.ANTIALIAS) y = self.model.predict({'image': im}) # coordinates are xywh normalized box = xywh2xyxy(y['coordinates'] * [[w, h, w, h]]) # xyxy pixels conf, cls = y['confidence'].max(1), y['confidence'].argmax(1).astype(np.float) y = np.concatenate((box, conf.reshape(-1, 1), cls.reshape(-1, 1)), 1) elif self.onnx: # ONNX im = im.cpu().numpy() # torch to numpy if self.dnn: # ONNX OpenCV DNN self.net.setInput(im) y = self.net.forward() else: # ONNX Runtime y = self.session.run([self.session.get_outputs()[0].name], {self.session.get_inputs()[0].name: im})[0] elif self.engine: # TensorRT assert im.shape == self.bindings['images'].shape, (im.shape, self.bindings['images'].shape) self.binding_addrs['images'] = int(im.data_ptr()) self.context.execute_v2(list(self.binding_addrs.values())) y = self.bindings['output'].data elif self.rknn_model: # TODO if post-process in model, then we can add code here. pass else: # TensorFlow model (TFLite, pb, saved_model) im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) if self.pb: y = self.frozen_func(x=self.tf.constant(im)).numpy() elif self.saved_model: y = self.model(im, training=False).numpy() elif self.tflite: input, output = self.input_details[0], self.output_details[0] int8 = input['dtype'] == np.uint8 # is TFLite quantized uint8 model if int8: scale, zero_point = input['quantization'] im = (im / scale + zero_point).astype(np.uint8) # de-scale self.interpreter.set_tensor(input['index'], im) self.interpreter.invoke() y = self.interpreter.get_tensor(output['index']) if int8: scale, zero_point = output['quantization'] y = (y.astype(np.float32) - zero_point) * scale # re-scale y[..., 0] *= w # x y[..., 1] *= h # y y[..., 2] *= w # w y[..., 3] *= h # h y = torch.tensor(y) if isinstance(y, np.ndarray) else y return (y, []) if val else y def warmup(self, imgsz=(1, 3, 640, 640), half=False): # Warmup model by running inference once if self.pt or self.engine or self.onnx: # warmup types if isinstance(self.device, torch.device) and self.device.type != 'cpu': # only warmup GPU models im = torch.zeros(*imgsz).to(self.device).type(torch.half if half else torch.float) # input image self.forward(im) # warmup class AutoShape(nn.Module): # YOLOv5 input-robust model wrapper for passing cv2/np/PIL/torch inputs. Includes preprocessing, inference and NMS conf = 0.25 # NMS confidence threshold iou = 0.45 # NMS IoU threshold agnostic = False # NMS class-agnostic multi_label = False # NMS multiple labels per box classes = None # (optional list) filter by class, i.e. = [0, 15, 16] for COCO persons, cats and dogs max_det = 1000 # maximum number of detections per image amp = False # Automatic Mixed Precision (AMP) inference def __init__(self, model): super().__init__() LOGGER.info('Adding AutoShape... ') copy_attr(self, model, include=('yaml', 'nc', 'hyp', 'names', 'stride', 'abc'), exclude=()) # copy attributes self.dmb = isinstance(model, DetectMultiBackend) # DetectMultiBackend() instance self.pt = not self.dmb or model.pt # PyTorch model self.model = model.eval() def _apply(self, fn): # Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers self = super()._apply(fn) if self.pt: m = self.model.model.model[-1] if self.dmb else self.model.model[-1] # Detect() m.stride = fn(m.stride) m.grid = list(map(fn, m.grid)) if isinstance(m.anchor_grid, list): m.anchor_grid = list(map(fn, m.anchor_grid)) return self @torch.no_grad() def forward(self, imgs, size=640, augment=False, profile=False): # Inference from various sources. For height=640, width=1280, RGB images example inputs are: # file: imgs = 'data/images/zidane.jpg' # str or PosixPath # URI: = 'https://ultralytics.com/images/zidane.jpg' # OpenCV: = cv2.imread('image.jpg')[:,:,::-1] # HWC BGR to RGB x(640,1280,3) # PIL: = Image.open('image.jpg') or ImageGrab.grab() # HWC x(640,1280,3) # numpy: = np.zeros((640,1280,3)) # HWC # torch: = torch.zeros(16,3,320,640) # BCHW (scaled to size=640, 0-1 values) # multiple: = [Image.open('image1.jpg'), Image.open('image2.jpg'), ...] # list of images t = [time_sync()] p = next(self.model.parameters()) if self.pt else torch.zeros(1) # for device and type autocast = self.amp and (p.device.type != 'cpu') # Automatic Mixed Precision (AMP) inference if isinstance(imgs, torch.Tensor): # torch with amp.autocast(enabled=autocast): return self.model(imgs.to(p.device).type_as(p), augment, profile) # inference # Pre-process n, imgs = (len(imgs), imgs) if isinstance(imgs, list) else (1, [imgs]) # number of images, list of images shape0, shape1, files = [], [], [] # image and inference shapes, filenames for i, im in enumerate(imgs): f = f'image{i}' # filename if isinstance(im, (str, Path)): # filename or uri im, f = Image.open(requests.get(im, stream=True).raw if str(im).startswith('http') else im), im im = np.asarray(exif_transpose(im)) elif isinstance(im, Image.Image): # PIL Image im, f = np.asarray(exif_transpose(im)), getattr(im, 'filename', f) or f files.append(Path(f).with_suffix('.jpg').name) if im.shape[0] < 5: # image in CHW im = im.transpose((1, 2, 0)) # reverse dataloader .transpose(2, 0, 1) im = im[..., :3] if im.ndim == 3 else np.tile(im[..., None], 3) # enforce 3ch input s = im.shape[:2] # HWC shape0.append(s) # image shape g = (size / max(s)) # gain shape1.append([y * g for y in s]) imgs[i] = im if im.data.contiguous else np.ascontiguousarray(im) # update shape1 = [make_divisible(x, self.stride) for x in np.stack(shape1, 0).max(0)] # inference shape x = [letterbox(im, new_shape=shape1 if self.pt else size, auto=False)[0] for im in imgs] # pad x = np.stack(x, 0) if n > 1 else x[0][None] # stack x = np.ascontiguousarray(x.transpose((0, 3, 1, 2))) # BHWC to BCHW x = torch.from_numpy(x).to(p.device).type_as(p) / 255 # uint8 to fp16/32 t.append(time_sync()) with amp.autocast(enabled=autocast): # Inference y = self.model(x, augment, profile) # forward t.append(time_sync()) # Post-process y = non_max_suppression(y if self.dmb else y[0], self.conf, iou_thres=self.iou, classes=self.classes, agnostic=self.agnostic, multi_label=self.multi_label, max_det=self.max_det) # NMS for i in range(n): scale_coords(shape1, y[i][:, :4], shape0[i]) t.append(time_sync()) return Detections(imgs, y, files, t, self.names, x.shape) class Detections: # YOLOv5 detections class for inference results def __init__(self, imgs, pred, files, times=(0, 0, 0, 0), names=None, shape=None): super().__init__() d = pred[0].device # device gn = [torch.tensor([*(im.shape[i] for i in [1, 0, 1, 0]), 1, 1], device=d) for im in imgs] # normalizations self.imgs = imgs # list of images as numpy arrays self.pred = pred # list of tensors pred[0] = (xyxy, conf, cls) self.names = names # class names self.files = files # image filenames self.times = times # profiling times self.xyxy = pred # xyxy pixels self.xywh = [xyxy2xywh(x) for x in pred] # xywh pixels self.xyxyn = [x / g for x, g in zip(self.xyxy, gn)] # xyxy normalized self.xywhn = [x / g for x, g in zip(self.xywh, gn)] # xywh normalized self.n = len(self.pred) # number of images (batch size) self.t = tuple((times[i + 1] - times[i]) * 1000 / self.n for i in range(3)) # timestamps (ms) self.s = shape # inference BCHW shape def display(self, pprint=False, show=False, save=False, crop=False, render=False, save_dir=Path('')): crops = [] for i, (im, pred) in enumerate(zip(self.imgs, self.pred)): s = f'image {i + 1}/{len(self.pred)}: {im.shape[0]}x{im.shape[1]} ' # string if pred.shape[0]: for c in pred[:, -1].unique(): n = (pred[:, -1] == c).sum() # detections per class s += f"{n} {self.names[int(c)]}{'s' * (n > 1)}, " # add to string if show or save or render or crop: annotator = Annotator(im, example=str(self.names)) for *box, conf, cls in reversed(pred): # xyxy, confidence, class label = f'{self.names[int(cls)]} {conf:.2f}' if crop: file = save_dir / 'crops' / self.names[int(cls)] / self.files[i] if save else None crops.append({'box': box, 'conf': conf, 'cls': cls, 'label': label, 'im': save_one_box(box, im, file=file, save=save)}) else: # all others annotator.box_label(box, label, color=colors(cls)) im = annotator.im else: s += '(no detections)' im = Image.fromarray(im.astype(np.uint8)) if isinstance(im, np.ndarray) else im # from np if pprint: LOGGER.info(s.rstrip(', ')) if show: im.show(self.files[i]) # show if save: f = self.files[i] im.save(save_dir / f) # save if i == self.n - 1: LOGGER.info(f"Saved {self.n} image{'s' * (self.n > 1)} to {colorstr('bold', save_dir)}") if render: self.imgs[i] = np.asarray(im) if crop: if save: LOGGER.info(f'Saved results to {save_dir}\n') return crops def print(self): self.display(pprint=True) # print results LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {tuple(self.s)}' % self.t) def show(self): self.display(show=True) # show results def save(self, save_dir='runs/detect/exp'): save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) # increment save_dir self.display(save=True, save_dir=save_dir) # save results def crop(self, save=True, save_dir='runs/detect/exp'): save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) if save else None return self.display(crop=True, save=save, save_dir=save_dir) # crop results def render(self): self.display(render=True) # render results return self.imgs def pandas(self): # return detections as pandas DataFrames, i.e. print(results.pandas().xyxy[0]) new = copy(self) # return copy ca = 'xmin', 'ymin', 'xmax', 'ymax', 'confidence', 'class', 'name' # xyxy columns cb = 'xcenter', 'ycenter', 'width', 'height', 'confidence', 'class', 'name' # xywh columns for k, c in zip(['xyxy', 'xyxyn', 'xywh', 'xywhn'], [ca, ca, cb, cb]): a = [[x[:5] + [int(x[5]), self.names[int(x[5])]] for x in x.tolist()] for x in getattr(self, k)] # update setattr(new, k, [pd.DataFrame(x, columns=c) for x in a]) return new def tolist(self): # return a list of Detections objects, i.e. 'for result in results.tolist():' r = range(self.n) # iterable x = [Detections([self.imgs[i]], [self.pred[i]], [self.files[i]], self.times, self.names, self.s) for i in r] # for d in x: # for k in ['imgs', 'pred', 'xyxy', 'xyxyn', 'xywh', 'xywhn']: # setattr(d, k, getattr(d, k)[0]) # pop out of list return x def __len__(self): return self.n class Classify(nn.Module): # Classification head, i.e. x(b,c1,20,20) to x(b,c2) def __init__(self, c1, c2, k=1, s=1, p=None, g=1): # ch_in, ch_out, kernel, stride, padding, groups super().__init__() self.aap = nn.AdaptiveAvgPool2d(1) # to x(b,c1,1,1) self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g) # to x(b,c2,1,1) self.flat = nn.Flatten() def forward(self, x): z = torch.cat([self.aap(y) for y in (x if isinstance(x, list) else [x])], 1) # cat if list return self.flat(self.conv(z)) # flatten to x(b,c2) ================================================ FILE: models/common_rk_plug_in.py ================================================ # This file contains modules common to various models import torch import torch.nn as nn from models.common import Conv class surrogate_focus(nn.Module): # surrogate_focus wh information into c-space def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups super(surrogate_focus, self).__init__() self.conv = Conv(c1 * 4, c2, k, s, p, g, act) with torch.no_grad(): self.conv1 = nn.Conv2d(3, 3, (2, 2), groups=3, bias=False, stride=(2, 2)) self.conv1.weight[:, :, 0, 0] = 1 self.conv1.weight[:, :, 0, 1] = 0 self.conv1.weight[:, :, 1, 0] = 0 self.conv1.weight[:, :, 1, 1] = 0 self.conv2 = nn.Conv2d(3, 3, (2, 2), groups=3, bias=False, stride=(2, 2)) self.conv2.weight[:, :, 0, 0] = 0 self.conv2.weight[:, :, 0, 1] = 0 self.conv2.weight[:, :, 1, 0] = 1 self.conv2.weight[:, :, 1, 1] = 0 self.conv3 = nn.Conv2d(3, 3, (2, 2), groups=3, bias=False, stride=(2, 2)) self.conv3.weight[:, :, 0, 0] = 0 self.conv3.weight[:, :, 0, 1] = 1 self.conv3.weight[:, :, 1, 0] = 0 self.conv3.weight[:, :, 1, 1] = 0 self.conv4 = nn.Conv2d(3, 3, (2, 2), groups=3, bias=False, stride=(2, 2)) self.conv4.weight[:, :, 0, 0] = 0 self.conv4.weight[:, :, 0, 1] = 0 self.conv4.weight[:, :, 1, 0] = 0 self.conv4.weight[:, :, 1, 1] = 1 def forward(self, x): # x(b,c,w,h) -> y(b,4c,w/2,h/2) return self.conv(torch.cat([self.conv1(x), self.conv2(x), self.conv3(x), self.conv4(x)], 1)) class preprocess_conv_layer(nn.Module): """docstring for preprocess_conv_layer""" # input_module 为输入模型，即为想要导出模型 # mean_value 的值可以是 [m1, m2, m3] 或 常数m # std_value 的值可以是 [s1, s2, s3] 或 常数s # BGR2RGB的操作默认为首先执行，既替代的原有操作顺序为 # BGR2RGB -> minus mean -> minus std (与rknn config 设置保持一致) -> nhwc2nchw # # 使用示例-伪代码： # from add_preprocess_conv_layer import preprocess_conv_layer # model_A = create_model() # model_output = preprocess_co_nv_layer(model_A, mean_value, std_value, BGR2RGB) # onnx_export(model_output) # # 量化时： # rknn.config的中 channel_mean_value 、reorder_channel 均不赋值。 # # 部署代码： # rknn_input 的属性 # pass_through = 1 # # 另外： # 由于加入permute操作，c端输入为opencv mat(hwc格式)即可，无需在外部将hwc改成chw格式。 # def __init__(self, input_module, mean_value, std_value, BGR2RGB=False): super(preprocess_conv_layer, self).__init__() if isinstance(mean_value, int): mean_value = [mean_value for i in range(3)] if isinstance(std_value, int): std_value = [std_value for i in range(3)] assert len(mean_value) <= 3, 'mean_value should be int, or list with 3 element' assert len(std_value) <= 3, 'std_value should be int, or list with 3 element' self.input_module = input_module with torch.no_grad(): self.conv1 = nn.Conv2d(3, 3, (1, 1), groups=1, bias=True, stride=(1, 1)) if BGR2RGB is False: self.conv1.weight[:, :, :, :] = 0 self.conv1.weight[0, 0, :, :] = 1/std_value[0] self.conv1.weight[1, 1, :, :] = 1/std_value[1] self.conv1.weight[2, 2, :, :] = 1/std_value[2] elif BGR2RGB is True: self.conv1.weight[:, :, :, :] = 0 self.conv1.weight[0, 2, :, :] = 1/std_value[0] self.conv1.weight[1, 1, :, :] = 1/std_value[1] self.conv1.weight[2, 0, :, :] = 1/std_value[2] self.conv1.bias[0] = -mean_value[0]/std_value[0] self.conv1.bias[1] = -mean_value[1]/std_value[1] self.conv1.bias[2] = -mean_value[2]/std_value[2] self.conv1.eval() def forward(self, x): x = x.permute(0, 3, 1, 2) # NHWC -> NCHW, apply for rknn_pass_through x = self.conv1(x) return self.input_module(x) ================================================ FILE: models/experimental.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Experimental modules """ import math import numpy as np import torch import torch.nn as nn from models.common import Conv from utils.downloads import attempt_download class CrossConv(nn.Module): # Cross Convolution Downsample def __init__(self, c1, c2, k=3, s=1, g=1, e=1.0, shortcut=False): # ch_in, ch_out, kernel, stride, groups, expansion, shortcut super().__init__() c_ = int(c2 * e) # hidden channels self.cv1 = Conv(c1, c_, (1, k), (1, s)) self.cv2 = Conv(c_, c2, (k, 1), (s, 1), g=g) self.add = shortcut and c1 == c2 def forward(self, x): return x + self.cv2(self.cv1(x)) if self.add else self.cv2(self.cv1(x)) class Sum(nn.Module): # Weighted sum of 2 or more layers https://arxiv.org/abs/1911.09070 def __init__(self, n, weight=False): # n: number of inputs super().__init__() self.weight = weight # apply weights boolean self.iter = range(n - 1) # iter object if weight: self.w = nn.Parameter(-torch.arange(1.0, n) / 2, requires_grad=True) # layer weights def forward(self, x): y = x[0] # no weight if self.weight: w = torch.sigmoid(self.w) * 2 for i in self.iter: y = y + x[i + 1] * w[i] else: for i in self.iter: y = y + x[i + 1] return y class MixConv2d(nn.Module): # Mixed Depth-wise Conv https://arxiv.org/abs/1907.09595 def __init__(self, c1, c2, k=(1, 3), s=1, equal_ch=True): # ch_in, ch_out, kernel, stride, ch_strategy super().__init__() n = len(k) # number of convolutions if equal_ch: # equal c_ per group i = torch.linspace(0, n - 1E-6, c2).floor() # c2 indices c_ = [(i == g).sum() for g in range(n)] # intermediate channels else: # equal weight.numel() per group b = [c2] + [0] * n a = np.eye(n + 1, n, k=-1) a -= np.roll(a, 1, axis=1) a *= np.array(k) ** 2 a[0] = 1 c_ = np.linalg.lstsq(a, b, rcond=None)[0].round() # solve for equal weight indices, ax = b self.m = nn.ModuleList( [nn.Conv2d(c1, int(c_), k, s, k // 2, groups=math.gcd(c1, int(c_)), bias=False) for k, c_ in zip(k, c_)]) self.bn = nn.BatchNorm2d(c2) self.act = nn.ReLU() def forward(self, x): return self.act(self.bn(torch.cat([m(x) for m in self.m], 1))) class Ensemble(nn.ModuleList): # Ensemble of models def __init__(self): super().__init__() def forward(self, x, augment=False, profile=False, visualize=False): y = [] for module in self: y.append(module(x, augment, profile, visualize)[0]) # y = torch.stack(y).max(0)[0] # max ensemble # y = torch.stack(y).mean(0) # mean ensemble y = torch.cat(y, 1) # nms ensemble return y, None # inference, train output def attempt_load(weights, map_location=None, inplace=True, fuse=True): from models.yolo import Detect, Model # Loads an ensemble of models weights=[a,b,c] or a single model weights=[a] or weights=a model = Ensemble() for w in weights if isinstance(weights, list) else [weights]: ckpt = torch.load(attempt_download(w), map_location=map_location) # load if fuse: model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().fuse().eval()) # FP32 model else: model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().eval()) # without layer fuse # Compatibility updates for m in model.modules(): if type(m) in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU, Detect, Model]: m.inplace = inplace # pytorch 1.7.0 compatibility if type(m) is Detect: if not isinstance(m.anchor_grid, list): # new Detect Layer compatibility delattr(m, 'anchor_grid') setattr(m, 'anchor_grid', [torch.zeros(1)] * m.nl) elif type(m) is Conv: m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility if len(model) == 1: return model[-1] # return model else: print(f'Ensemble created with {weights}\n') for k in ['names']: setattr(model, k, getattr(model[-1], k)) model.stride = model[torch.argmax(torch.tensor([m.stride.max() for m in model])).int()].stride # max stride return model # return ensemble ================================================ FILE: models/hub/anchors.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Default anchors for COCO data # P5 ------------------------------------------------------------------------------------------------------------------- # P5-640: anchors_p5_640: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # P6 ------------------------------------------------------------------------------------------------------------------- # P6-640: thr=0.25: 0.9964 BPR, 5.54 anchors past thr, n=12, img_size=640, metric_all=0.281/0.716-mean/best, past_thr=0.469-mean: 9,11, 21,19, 17,41, 43,32, 39,70, 86,64, 65,131, 134,130, 120,265, 282,180, 247,354, 512,387 anchors_p6_640: - [9,11, 21,19, 17,41] # P3/8 - [43,32, 39,70, 86,64] # P4/16 - [65,131, 134,130, 120,265] # P5/32 - [282,180, 247,354, 512,387] # P6/64 # P6-1280: thr=0.25: 0.9950 BPR, 5.55 anchors past thr, n=12, img_size=1280, metric_all=0.281/0.714-mean/best, past_thr=0.468-mean: 19,27, 44,40, 38,94, 96,68, 86,152, 180,137, 140,301, 303,264, 238,542, 436,615, 739,380, 925,792 anchors_p6_1280: - [19,27, 44,40, 38,94] # P3/8 - [96,68, 86,152, 180,137] # P4/16 - [140,301, 303,264, 238,542] # P5/32 - [436,615, 739,380, 925,792] # P6/64 # P6-1920: thr=0.25: 0.9950 BPR, 5.55 anchors past thr, n=12, img_size=1920, metric_all=0.281/0.714-mean/best, past_thr=0.468-mean: 28,41, 67,59, 57,141, 144,103, 129,227, 270,205, 209,452, 455,396, 358,812, 653,922, 1109,570, 1387,1187 anchors_p6_1920: - [28,41, 67,59, 57,141] # P3/8 - [144,103, 129,227, 270,205] # P4/16 - [209,452, 455,396, 358,812] # P5/32 - [653,922, 1109,570, 1387,1187] # P6/64 # P7 ------------------------------------------------------------------------------------------------------------------- # P7-640: thr=0.25: 0.9962 BPR, 6.76 anchors past thr, n=15, img_size=640, metric_all=0.275/0.733-mean/best, past_thr=0.466-mean: 11,11, 13,30, 29,20, 30,46, 61,38, 39,92, 78,80, 146,66, 79,163, 149,150, 321,143, 157,303, 257,402, 359,290, 524,372 anchors_p7_640: - [11,11, 13,30, 29,20] # P3/8 - [30,46, 61,38, 39,92] # P4/16 - [78,80, 146,66, 79,163] # P5/32 - [149,150, 321,143, 157,303] # P6/64 - [257,402, 359,290, 524,372] # P7/128 # P7-1280: thr=0.25: 0.9968 BPR, 6.71 anchors past thr, n=15, img_size=1280, metric_all=0.273/0.732-mean/best, past_thr=0.463-mean: 19,22, 54,36, 32,77, 70,83, 138,71, 75,173, 165,159, 148,334, 375,151, 334,317, 251,626, 499,474, 750,326, 534,814, 1079,818 anchors_p7_1280: - [19,22, 54,36, 32,77] # P3/8 - [70,83, 138,71, 75,173] # P4/16 - [165,159, 148,334, 375,151] # P5/32 - [334,317, 251,626, 499,474] # P6/64 - [750,326, 534,814, 1079,818] # P7/128 # P7-1920: thr=0.25: 0.9968 BPR, 6.71 anchors past thr, n=15, img_size=1920, metric_all=0.273/0.732-mean/best, past_thr=0.463-mean: 29,34, 81,55, 47,115, 105,124, 207,107, 113,259, 247,238, 222,500, 563,227, 501,476, 376,939, 749,711, 1126,489, 801,1222, 1618,1227 anchors_p7_1920: - [29,34, 81,55, 47,115] # P3/8 - [105,124, 207,107, 113,259] # P4/16 - [247,238, 222,500, 563,227] # P5/32 - [501,476, 376,939, 749,711] # P6/64 - [1126,489, 801,1222, 1618,1227] # P7/128 ================================================ FILE: models/hub/yolov3-spp.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # darknet53 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [32, 3, 1]], # 0 [-1, 1, Conv, [64, 3, 2]], # 1-P1/2 [-1, 1, Bottleneck, [64]], [-1, 1, Conv, [128, 3, 2]], # 3-P2/4 [-1, 2, Bottleneck, [128]], [-1, 1, Conv, [256, 3, 2]], # 5-P3/8 [-1, 8, Bottleneck, [256]], [-1, 1, Conv, [512, 3, 2]], # 7-P4/16 [-1, 8, Bottleneck, [512]], [-1, 1, Conv, [1024, 3, 2]], # 9-P5/32 [-1, 4, Bottleneck, [1024]], # 10 ] # YOLOv3-SPP head head: [[-1, 1, Bottleneck, [1024, False]], [-1, 1, SPP, [512, [5, 9, 13]]], [-1, 1, Conv, [1024, 3, 1]], [-1, 1, Conv, [512, 1, 1]], [-1, 1, Conv, [1024, 3, 1]], # 15 (P5/32-large) [-2, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P4 [-1, 1, Bottleneck, [512, False]], [-1, 1, Bottleneck, [512, False]], [-1, 1, Conv, [256, 1, 1]], [-1, 1, Conv, [512, 3, 1]], # 22 (P4/16-medium) [-2, 1, Conv, [128, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P3 [-1, 1, Bottleneck, [256, False]], [-1, 2, Bottleneck, [256, False]], # 27 (P3/8-small) [[27, 22, 15], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/hub/yolov3-tiny.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [10,14, 23,27, 37,58] # P4/16 - [81,82, 135,169, 344,319] # P5/32 # YOLOv3-tiny backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [16, 3, 1]], # 0 [-1, 1, nn.MaxPool2d, [2, 2, 0]], # 1-P1/2 [-1, 1, Conv, [32, 3, 1]], [-1, 1, nn.MaxPool2d, [2, 2, 0]], # 3-P2/4 [-1, 1, Conv, [64, 3, 1]], [-1, 1, nn.MaxPool2d, [2, 2, 0]], # 5-P3/8 [-1, 1, Conv, [128, 3, 1]], [-1, 1, nn.MaxPool2d, [2, 2, 0]], # 7-P4/16 [-1, 1, Conv, [256, 3, 1]], [-1, 1, nn.MaxPool2d, [2, 2, 0]], # 9-P5/32 [-1, 1, Conv, [512, 3, 1]], [-1, 1, nn.ZeroPad2d, [[0, 1, 0, 1]]], # 11 [-1, 1, nn.MaxPool2d, [2, 1, 0]], # 12 ] # YOLOv3-tiny head head: [[-1, 1, Conv, [1024, 3, 1]], [-1, 1, Conv, [256, 1, 1]], [-1, 1, Conv, [512, 3, 1]], # 15 (P5/32-large) [-2, 1, Conv, [128, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P4 [-1, 1, Conv, [256, 3, 1]], # 19 (P4/16-medium) [[19, 15], 1, Detect, [nc, anchors]], # Detect(P4, P5) ] ================================================ FILE: models/hub/yolov3.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # darknet53 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [32, 3, 1]], # 0 [-1, 1, Conv, [64, 3, 2]], # 1-P1/2 [-1, 1, Bottleneck, [64]], [-1, 1, Conv, [128, 3, 2]], # 3-P2/4 [-1, 2, Bottleneck, [128]], [-1, 1, Conv, [256, 3, 2]], # 5-P3/8 [-1, 8, Bottleneck, [256]], [-1, 1, Conv, [512, 3, 2]], # 7-P4/16 [-1, 8, Bottleneck, [512]], [-1, 1, Conv, [1024, 3, 2]], # 9-P5/32 [-1, 4, Bottleneck, [1024]], # 10 ] # YOLOv3 head head: [[-1, 1, Bottleneck, [1024, False]], [-1, 1, Conv, [512, [1, 1]]], [-1, 1, Conv, [1024, 3, 1]], [-1, 1, Conv, [512, 1, 1]], [-1, 1, Conv, [1024, 3, 1]], # 15 (P5/32-large) [-2, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P4 [-1, 1, Bottleneck, [512, False]], [-1, 1, Bottleneck, [512, False]], [-1, 1, Conv, [256, 1, 1]], [-1, 1, Conv, [512, 3, 1]], # 22 (P4/16-medium) [-2, 1, Conv, [128, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P3 [-1, 1, Bottleneck, [256, False]], [-1, 2, Bottleneck, [256, False]], # 27 (P3/8-small) [[27, 22, 15], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/hub/yolov5-bifpn.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 BiFPN head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14, 6], 1, Concat, [1]], # cat P4 <--- BiFPN change [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/hub/yolov5-fpn.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 FPN head head: [[-1, 3, C3, [1024, False]], # 10 (P5/32-large) [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 1, Conv, [512, 1, 1]], [-1, 3, C3, [512, False]], # 14 (P4/16-medium) [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 1, Conv, [256, 1, 1]], [-1, 3, C3, [256, False]], # 18 (P3/8-small) [[18, 14, 10], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/hub/yolov5-p2.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: 3 # auto-anchor evolves 3 anchors per P output layer # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [128, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 2], 1, Concat, [1]], # cat backbone P2 [-1, 1, C3, [128, False]], # 21 (P2/4-xsmall) [-1, 1, Conv, [128, 3, 2]], [[-1, 18], 1, Concat, [1]], # cat head P3 [-1, 3, C3, [256, False]], # 24 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 27 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 30 (P5/32-large) [[21, 24, 27, 30], 1, Detect, [nc, anchors]], # Detect(P2, P3, P4, P5) ] ================================================ FILE: models/hub/yolov5-p6.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: 3 # auto-anchor 3 anchors per P output layer # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [768, 3, 2]], # 7-P5/32 [-1, 3, C3, [768]], [-1, 1, Conv, [1024, 3, 2]], # 9-P6/64 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 11 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [768, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P5 [-1, 3, C3, [768, False]], # 15 [-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 19 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 23 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 20], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 26 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 16], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [768, False]], # 29 (P5/32-large) [-1, 1, Conv, [768, 3, 2]], [[-1, 12], 1, Concat, [1]], # cat head P6 [-1, 3, C3, [1024, False]], # 32 (P6/64-xlarge) [[23, 26, 29, 32], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5, P6) ] ================================================ FILE: models/hub/yolov5-p7.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: 3 # auto-anchor 3 anchors per P output layer # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [768, 3, 2]], # 7-P5/32 [-1, 3, C3, [768]], [-1, 1, Conv, [1024, 3, 2]], # 9-P6/64 [-1, 3, C3, [1024]], [-1, 1, Conv, [1280, 3, 2]], # 11-P7/128 [-1, 3, C3, [1280]], [-1, 1, SPPF, [1280, 5]], # 13 ] # YOLOv5 head head: [[-1, 1, Conv, [1024, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 10], 1, Concat, [1]], # cat backbone P6 [-1, 3, C3, [1024, False]], # 17 [-1, 1, Conv, [768, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P5 [-1, 3, C3, [768, False]], # 21 [-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 25 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 29 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 26], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 32 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 22], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [768, False]], # 35 (P5/32-large) [-1, 1, Conv, [768, 3, 2]], [[-1, 18], 1, Concat, [1]], # cat head P6 [-1, 3, C3, [1024, False]], # 38 (P6/64-xlarge) [-1, 1, Conv, [1024, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P7 [-1, 3, C3, [1280, False]], # 41 (P7/128-xxlarge) [[29, 32, 35, 38, 41], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5, P6, P7) ] ================================================ FILE: models/hub/yolov5-panet.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 PANet head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/hub/yolov5l6.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [19,27, 44,40, 38,94] # P3/8 - [96,68, 86,152, 180,137] # P4/16 - [140,301, 303,264, 238,542] # P5/32 - [436,615, 739,380, 925,792] # P6/64 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [768, 3, 2]], # 7-P5/32 [-1, 3, C3, [768]], [-1, 1, Conv, [1024, 3, 2]], # 9-P6/64 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 11 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [768, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P5 [-1, 3, C3, [768, False]], # 15 [-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 19 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 23 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 20], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 26 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 16], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [768, False]], # 29 (P5/32-large) [-1, 1, Conv, [768, 3, 2]], [[-1, 12], 1, Concat, [1]], # cat head P6 [-1, 3, C3, [1024, False]], # 32 (P6/64-xlarge) [[23, 26, 29, 32], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5, P6) ] ================================================ FILE: models/hub/yolov5m6.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.67 # model depth multiple width_multiple: 0.75 # layer channel multiple anchors: - [19,27, 44,40, 38,94] # P3/8 - [96,68, 86,152, 180,137] # P4/16 - [140,301, 303,264, 238,542] # P5/32 - [436,615, 739,380, 925,792] # P6/64 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [768, 3, 2]], # 7-P5/32 [-1, 3, C3, [768]], [-1, 1, Conv, [1024, 3, 2]], # 9-P6/64 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 11 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [768, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P5 [-1, 3, C3, [768, False]], # 15 [-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 19 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 23 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 20], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 26 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 16], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [768, False]], # 29 (P5/32-large) [-1, 1, Conv, [768, 3, 2]], [[-1, 12], 1, Concat, [1]], # cat head P6 [-1, 3, C3, [1024, False]], # 32 (P6/64-xlarge) [[23, 26, 29, 32], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5, P6) ] ================================================ FILE: models/hub/yolov5n6.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.33 # model depth multiple width_multiple: 0.25 # layer channel multiple anchors: - [19,27, 44,40, 38,94] # P3/8 - [96,68, 86,152, 180,137] # P4/16 - [140,301, 303,264, 238,542] # P5/32 - [436,615, 739,380, 925,792] # P6/64 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [768, 3, 2]], # 7-P5/32 [-1, 3, C3, [768]], [-1, 1, Conv, [1024, 3, 2]], # 9-P6/64 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 11 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [768, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P5 [-1, 3, C3, [768, False]], # 15 [-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 19 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 23 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 20], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 26 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 16], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [768, False]], # 29 (P5/32-large) [-1, 1, Conv, [768, 3, 2]], [[-1, 12], 1, Concat, [1]], # cat head P6 [-1, 3, C3, [1024, False]], # 32 (P6/64-xlarge) [[23, 26, 29, 32], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5, P6) ] ================================================ FILE: models/hub/yolov5s-ghost.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.33 # model depth multiple width_multiple: 0.50 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, GhostConv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3Ghost, [128]], [-1, 1, GhostConv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3Ghost, [256]], [-1, 1, GhostConv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3Ghost, [512]], [-1, 1, GhostConv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3Ghost, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, GhostConv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3Ghost, [512, False]], # 13 [-1, 1, GhostConv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3Ghost, [256, False]], # 17 (P3/8-small) [-1, 1, GhostConv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3Ghost, [512, False]], # 20 (P4/16-medium) [-1, 1, GhostConv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3Ghost, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/hub/yolov5s-transformer.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.33 # model depth multiple width_multiple: 0.50 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3TR, [1024]], # 9 <--- C3TR() Transformer module [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/hub/yolov5s6.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.33 # model depth multiple width_multiple: 0.50 # layer channel multiple anchors: - [19,27, 44,40, 38,94] # P3/8 - [96,68, 86,152, 180,137] # P4/16 - [140,301, 303,264, 238,542] # P5/32 - [436,615, 739,380, 925,792] # P6/64 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [768, 3, 2]], # 7-P5/32 [-1, 3, C3, [768]], [-1, 1, Conv, [1024, 3, 2]], # 9-P6/64 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 11 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [768, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P5 [-1, 3, C3, [768, False]], # 15 [-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 19 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 23 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 20], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 26 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 16], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [768, False]], # 29 (P5/32-large) [-1, 1, Conv, [768, 3, 2]], [[-1, 12], 1, Concat, [1]], # cat head P6 [-1, 3, C3, [1024, False]], # 32 (P6/64-xlarge) [[23, 26, 29, 32], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5, P6) ] ================================================ FILE: models/hub/yolov5x6.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.33 # model depth multiple width_multiple: 1.25 # layer channel multiple anchors: - [19,27, 44,40, 38,94] # P3/8 - [96,68, 86,152, 180,137] # P4/16 - [140,301, 303,264, 238,542] # P5/32 - [436,615, 739,380, 925,792] # P6/64 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [768, 3, 2]], # 7-P5/32 [-1, 3, C3, [768]], [-1, 1, Conv, [1024, 3, 2]], # 9-P6/64 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 11 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [768, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 8], 1, Concat, [1]], # cat backbone P5 [-1, 3, C3, [768, False]], # 15 [-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 19 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 23 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 20], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 26 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 16], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [768, False]], # 29 (P5/32-large) [-1, 1, Conv, [768, 3, 2]], [[-1, 12], 1, Concat, [1]], # cat head P6 [-1, 3, C3, [1024, False]], # 32 (P6/64-xlarge) [[23, 26, 29, 32], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5, P6) ] ================================================ FILE: models/tf.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ TensorFlow, Keras and TFLite versions of YOLOv5 Authored by https://github.com/zldrobit in PR https://github.com/ultralytics/yolov5/pull/1127 Usage: $ python models/tf.py --weights yolov5s.pt Export: $ python path/to/export.py --weights yolov5s.pt --include saved_model pb tflite tfjs """ import argparse import sys from copy import deepcopy from pathlib import Path FILE = Path(__file__).resolve() ROOT = FILE.parents[1] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH # ROOT = ROOT.relative_to(Path.cwd()) # relative import numpy as np import tensorflow as tf import torch import torch.nn as nn from tensorflow import keras from models.common import C3, SPP, SPPF, Bottleneck, BottleneckCSP, Concat, Conv, DWConv, Focus, autopad from models.experimental import CrossConv, MixConv2d, attempt_load from models.yolo import Detect from utils.activations import SiLU from utils.general import LOGGER, make_divisible, print_args class TFBN(keras.layers.Layer): # TensorFlow BatchNormalization wrapper def __init__(self, w=None): super().__init__() self.bn = keras.layers.BatchNormalization( beta_initializer=keras.initializers.Constant(w.bias.numpy()), gamma_initializer=keras.initializers.Constant(w.weight.numpy()), moving_mean_initializer=keras.initializers.Constant(w.running_mean.numpy()), moving_variance_initializer=keras.initializers.Constant(w.running_var.numpy()), epsilon=w.eps) def call(self, inputs): return self.bn(inputs) class TFPad(keras.layers.Layer): def __init__(self, pad): super().__init__() self.pad = tf.constant([[0, 0], [pad, pad], [pad, pad], [0, 0]]) def call(self, inputs): return tf.pad(inputs, self.pad, mode='constant', constant_values=0) class TFConv(keras.layers.Layer): # Standard convolution def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True, w=None): # ch_in, ch_out, weights, kernel, stride, padding, groups super().__init__() assert g == 1, "TF v2.2 Conv2D does not support 'groups' argument" assert isinstance(k, int), "Convolution with multiple kernels are not allowed." # TensorFlow convolution padding is inconsistent with PyTorch (e.g. k=3 s=2 'SAME' padding) # see https://stackoverflow.com/questions/52975843/comparing-conv2d-with-padding-between-tensorflow-and-pytorch conv = keras.layers.Conv2D( c2, k, s, 'SAME' if s == 1 else 'VALID', use_bias=False if hasattr(w, 'bn') else True, kernel_initializer=keras.initializers.Constant(w.conv.weight.permute(2, 3, 1, 0).numpy()), bias_initializer='zeros' if hasattr(w, 'bn') else keras.initializers.Constant(w.conv.bias.numpy())) self.conv = conv if s == 1 else keras.Sequential([TFPad(autopad(k, p)), conv]) self.bn = TFBN(w.bn) if hasattr(w, 'bn') else tf.identity # YOLOv5 activations if isinstance(w.act, nn.LeakyReLU): self.act = (lambda x: keras.activations.relu(x, alpha=0.1)) if act else tf.identity elif isinstance(w.act, nn.Hardswish): self.act = (lambda x: x * tf.nn.relu6(x + 3) * 0.166666667) if act else tf.identity elif isinstance(w.act, (nn.SiLU, SiLU)): self.act = (lambda x: keras.activations.swish(x)) if act else tf.identity else: raise Exception(f'no matching TensorFlow activation found for {w.act}') def call(self, inputs): return self.act(self.bn(self.conv(inputs))) class TFFocus(keras.layers.Layer): # Focus wh information into c-space def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True, w=None): # ch_in, ch_out, kernel, stride, padding, groups super().__init__() self.conv = TFConv(c1 * 4, c2, k, s, p, g, act, w.conv) def call(self, inputs): # x(b,w,h,c) -> y(b,w/2,h/2,4c) # inputs = inputs / 255 # normalize 0-255 to 0-1 return self.conv(tf.concat([inputs[:, ::2, ::2, :], inputs[:, 1::2, ::2, :], inputs[:, ::2, 1::2, :], inputs[:, 1::2, 1::2, :]], 3)) class TFBottleneck(keras.layers.Layer): # Standard bottleneck def __init__(self, c1, c2, shortcut=True, g=1, e=0.5, w=None): # ch_in, ch_out, shortcut, groups, expansion super().__init__() c_ = int(c2 * e) # hidden channels self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1) self.cv2 = TFConv(c_, c2, 3, 1, g=g, w=w.cv2) self.add = shortcut and c1 == c2 def call(self, inputs): return inputs + self.cv2(self.cv1(inputs)) if self.add else self.cv2(self.cv1(inputs)) class TFConv2d(keras.layers.Layer): # Substitution for PyTorch nn.Conv2D def __init__(self, c1, c2, k, s=1, g=1, bias=True, w=None): super().__init__() assert g == 1, "TF v2.2 Conv2D does not support 'groups' argument" self.conv = keras.layers.Conv2D( c2, k, s, 'VALID', use_bias=bias, kernel_initializer=keras.initializers.Constant(w.weight.permute(2, 3, 1, 0).numpy()), bias_initializer=keras.initializers.Constant(w.bias.numpy()) if bias else None, ) def call(self, inputs): return self.conv(inputs) class TFBottleneckCSP(keras.layers.Layer): # CSP Bottleneck https://github.com/WongKinYiu/CrossStagePartialNetworks def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5, w=None): # ch_in, ch_out, number, shortcut, groups, expansion super().__init__() c_ = int(c2 * e) # hidden channels self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1) self.cv2 = TFConv2d(c1, c_, 1, 1, bias=False, w=w.cv2) self.cv3 = TFConv2d(c_, c_, 1, 1, bias=False, w=w.cv3) self.cv4 = TFConv(2 * c_, c2, 1, 1, w=w.cv4) self.bn = TFBN(w.bn) self.act = lambda x: keras.activations.relu(x, alpha=0.1) self.m = keras.Sequential([TFBottleneck(c_, c_, shortcut, g, e=1.0, w=w.m[j]) for j in range(n)]) def call(self, inputs): y1 = self.cv3(self.m(self.cv1(inputs))) y2 = self.cv2(inputs) return self.cv4(self.act(self.bn(tf.concat((y1, y2), axis=3)))) class TFC3(keras.layers.Layer): # CSP Bottleneck with 3 convolutions def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5, w=None): # ch_in, ch_out, number, shortcut, groups, expansion super().__init__() c_ = int(c2 * e) # hidden channels self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1) self.cv2 = TFConv(c1, c_, 1, 1, w=w.cv2) self.cv3 = TFConv(2 * c_, c2, 1, 1, w=w.cv3) self.m = keras.Sequential([TFBottleneck(c_, c_, shortcut, g, e=1.0, w=w.m[j]) for j in range(n)]) def call(self, inputs): return self.cv3(tf.concat((self.m(self.cv1(inputs)), self.cv2(inputs)), axis=3)) class TFSPP(keras.layers.Layer): # Spatial pyramid pooling layer used in YOLOv3-SPP def __init__(self, c1, c2, k=(5, 9, 13), w=None): super().__init__() c_ = c1 // 2 # hidden channels self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1) self.cv2 = TFConv(c_ * (len(k) + 1), c2, 1, 1, w=w.cv2) self.m = [keras.layers.MaxPool2D(pool_size=x, strides=1, padding='SAME') for x in k] def call(self, inputs): x = self.cv1(inputs) return self.cv2(tf.concat([x] + [m(x) for m in self.m], 3)) class TFSPPF(keras.layers.Layer): # Spatial pyramid pooling-Fast layer def __init__(self, c1, c2, k=5, w=None): super().__init__() c_ = c1 // 2 # hidden channels self.cv1 = TFConv(c1, c_, 1, 1, w=w.cv1) self.cv2 = TFConv(c_ * 4, c2, 1, 1, w=w.cv2) self.m = keras.layers.MaxPool2D(pool_size=k, strides=1, padding='SAME') def call(self, inputs): x = self.cv1(inputs) y1 = self.m(x) y2 = self.m(y1) return self.cv2(tf.concat([x, y1, y2, self.m(y2)], 3)) class TFDetect(keras.layers.Layer): def __init__(self, nc=80, anchors=(), ch=(), imgsz=(640, 640), w=None): # detection layer super().__init__() self.stride = tf.convert_to_tensor(w.stride.numpy(), dtype=tf.float32) self.nc = nc # number of classes self.no = nc + 5 # number of outputs per anchor self.nl = len(anchors) # number of detection layers self.na = len(anchors[0]) // 2 # number of anchors self.grid = [tf.zeros(1)] * self.nl # init grid self.anchors = tf.convert_to_tensor(w.anchors.numpy(), dtype=tf.float32) self.anchor_grid = tf.reshape(self.anchors * tf.reshape(self.stride, [self.nl, 1, 1]), [self.nl, 1, -1, 1, 2]) self.m = [TFConv2d(x, self.no * self.na, 1, w=w.m[i]) for i, x in enumerate(ch)] self.training = False # set to False after building model self.imgsz = imgsz for i in range(self.nl): ny, nx = self.imgsz[0] // self.stride[i], self.imgsz[1] // self.stride[i] self.grid[i] = self._make_grid(nx, ny) def call(self, inputs): z = [] # inference output x = [] for i in range(self.nl): x.append(self.m[i](inputs[i])) # x(bs,20,20,255) to x(bs,3,20,20,85) ny, nx = self.imgsz[0] // self.stride[i], self.imgsz[1] // self.stride[i] x[i] = tf.transpose(tf.reshape(x[i], [-1, ny * nx, self.na, self.no]), [0, 2, 1, 3]) if not self.training: # inference y = tf.sigmoid(x[i]) xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # Normalize xywh to 0-1 to reduce calibration error xy /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32) wh /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32) y = tf.concat([xy, wh, y[..., 4:]], -1) z.append(tf.reshape(y, [-1, self.na * ny * nx, self.no])) return x if self.training else (tf.concat(z, 1), x) @staticmethod def _make_grid(nx=20, ny=20): # yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)]) # return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float() xv, yv = tf.meshgrid(tf.range(nx), tf.range(ny)) return tf.cast(tf.reshape(tf.stack([xv, yv], 2), [1, 1, ny * nx, 2]), dtype=tf.float32) class TFUpsample(keras.layers.Layer): def __init__(self, size, scale_factor, mode, w=None): # warning: all arguments needed including 'w' super().__init__() assert scale_factor == 2, "scale_factor must be 2" self.upsample = lambda x: tf.image.resize(x, (x.shape[1] * 2, x.shape[2] * 2), method=mode) # self.upsample = keras.layers.UpSampling2D(size=scale_factor, interpolation=mode) # with default arguments: align_corners=False, half_pixel_centers=False # self.upsample = lambda x: tf.raw_ops.ResizeNearestNeighbor(images=x, # size=(x.shape[1] * 2, x.shape[2] * 2)) def call(self, inputs): return self.upsample(inputs) class TFConcat(keras.layers.Layer): def __init__(self, dimension=1, w=None): super().__init__() assert dimension == 1, "convert only NCHW to NHWC concat" self.d = 3 def call(self, inputs): return tf.concat(inputs, self.d) def parse_model(d, ch, model, imgsz): # model_dict, input_channels(3) LOGGER.info(f"\n{'':>3}{'from':>18}{'n':>3}{'params':>10} {'module':<40}{'arguments':<30}") anchors, nc, gd, gw = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple'] na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors # number of anchors no = na * (nc + 5) # number of outputs = anchors * (classes + 5) layers, save, c2 = [], [], ch[-1] # layers, savelist, ch out for i, (f, n, m, args) in enumerate(d['backbone'] + d['head']): # from, number, module, args m_str = m m = eval(m) if isinstance(m, str) else m # eval strings for j, a in enumerate(args): try: args[j] = eval(a) if isinstance(a, str) else a # eval strings except NameError: pass n = max(round(n * gd), 1) if n > 1 else n # depth gain if m in [nn.Conv2d, Conv, Bottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, BottleneckCSP, C3]: c1, c2 = ch[f], args[0] c2 = make_divisible(c2 * gw, 8) if c2 != no else c2 args = [c1, c2, *args[1:]] if m in [BottleneckCSP, C3]: args.insert(2, n) n = 1 elif m is nn.BatchNorm2d: args = [ch[f]] elif m is Concat: c2 = sum(ch[-1 if x == -1 else x + 1] for x in f) elif m is Detect: args.append([ch[x + 1] for x in f]) if isinstance(args[1], int): # number of anchors args[1] = [list(range(args[1] * 2))] * len(f) args.append(imgsz) else: c2 = ch[f] tf_m = eval('TF' + m_str.replace('nn.', '')) m_ = keras.Sequential([tf_m(*args, w=model.model[i][j]) for j in range(n)]) if n > 1 \ else tf_m(*args, w=model.model[i]) # module torch_m_ = nn.Sequential(*(m(*args) for _ in range(n))) if n > 1 else m(*args) # module t = str(m)[8:-2].replace('__main__.', '') # module type np = sum(x.numel() for x in torch_m_.parameters()) # number params m_.i, m_.f, m_.type, m_.np = i, f, t, np # attach index, 'from' index, type, number params LOGGER.info(f'{i:>3}{str(f):>18}{str(n):>3}{np:>10} {t:<40}{str(args):<30}') # print save.extend(x % i for x in ([f] if isinstance(f, int) else f) if x != -1) # append to savelist layers.append(m_) ch.append(c2) return keras.Sequential(layers), sorted(save) class TFModel: def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, model=None, imgsz=(640, 640)): # model, channels, classes super().__init__() if isinstance(cfg, dict): self.yaml = cfg # model dict else: # is *.yaml import yaml # for torch hub self.yaml_file = Path(cfg).name with open(cfg) as f: self.yaml = yaml.load(f, Loader=yaml.FullLoader) # model dict # Define model if nc and nc != self.yaml['nc']: LOGGER.info(f"Overriding {cfg} nc={self.yaml['nc']} with nc={nc}") self.yaml['nc'] = nc # override yaml value self.model, self.savelist = parse_model(deepcopy(self.yaml), ch=[ch], model=model, imgsz=imgsz) def predict(self, inputs, tf_nms=False, agnostic_nms=False, topk_per_class=100, topk_all=100, iou_thres=0.45, conf_thres=0.25): y = [] # outputs x = inputs for i, m in enumerate(self.model.layers): if m.f != -1: # if not from previous layer x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers x = m(x) # run y.append(x if m.i in self.savelist else None) # save output # Add TensorFlow NMS if tf_nms: boxes = self._xywh2xyxy(x[0][..., :4]) probs = x[0][:, :, 4:5] classes = x[0][:, :, 5:] scores = probs * classes if agnostic_nms: nms = AgnosticNMS()((boxes, classes, scores), topk_all, iou_thres, conf_thres) return nms, x[1] else: boxes = tf.expand_dims(boxes, 2) nms = tf.image.combined_non_max_suppression( boxes, scores, topk_per_class, topk_all, iou_thres, conf_thres, clip_boxes=False) return nms, x[1] return x[0] # output only first tensor [1,6300,85] = [xywh, conf, class0, class1, ...] # x = x[0][0] # [x(1,6300,85), ...] to x(6300,85) # xywh = x[..., :4] # x(6300,4) boxes # conf = x[..., 4:5] # x(6300,1) confidences # cls = tf.reshape(tf.cast(tf.argmax(x[..., 5:], axis=1), tf.float32), (-1, 1)) # x(6300,1) classes # return tf.concat([conf, cls, xywh], 1) @staticmethod def _xywh2xyxy(xywh): # Convert nx4 boxes from [x, y, w, h] to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right x, y, w, h = tf.split(xywh, num_or_size_splits=4, axis=-1) return tf.concat([x - w / 2, y - h / 2, x + w / 2, y + h / 2], axis=-1) class AgnosticNMS(keras.layers.Layer): # TF Agnostic NMS def call(self, input, topk_all, iou_thres, conf_thres): # wrap map_fn to avoid TypeSpec related error https://stackoverflow.com/a/65809989/3036450 return tf.map_fn(lambda x: self._nms(x, topk_all, iou_thres, conf_thres), input, fn_output_signature=(tf.float32, tf.float32, tf.float32, tf.int32), name='agnostic_nms') @staticmethod def _nms(x, topk_all=100, iou_thres=0.45, conf_thres=0.25): # agnostic NMS boxes, classes, scores = x class_inds = tf.cast(tf.argmax(classes, axis=-1), tf.float32) scores_inp = tf.reduce_max(scores, -1) selected_inds = tf.image.non_max_suppression( boxes, scores_inp, max_output_size=topk_all, iou_threshold=iou_thres, score_threshold=conf_thres) selected_boxes = tf.gather(boxes, selected_inds) padded_boxes = tf.pad(selected_boxes, paddings=[[0, topk_all - tf.shape(selected_boxes)[0]], [0, 0]], mode="CONSTANT", constant_values=0.0) selected_scores = tf.gather(scores_inp, selected_inds) padded_scores = tf.pad(selected_scores, paddings=[[0, topk_all - tf.shape(selected_boxes)[0]]], mode="CONSTANT", constant_values=-1.0) selected_classes = tf.gather(class_inds, selected_inds) padded_classes = tf.pad(selected_classes, paddings=[[0, topk_all - tf.shape(selected_boxes)[0]]], mode="CONSTANT", constant_values=-1.0) valid_detections = tf.shape(selected_inds)[0] return padded_boxes, padded_scores, padded_classes, valid_detections def representative_dataset_gen(dataset, ncalib=100): # Representative dataset generator for use with converter.representative_dataset, returns a generator of np arrays for n, (path, img, im0s, vid_cap, string) in enumerate(dataset): input = np.transpose(img, [1, 2, 0]) input = np.expand_dims(input, axis=0).astype(np.float32) input /= 255 yield [input] if n >= ncalib: break def run(weights=ROOT / 'yolov5s.pt', # weights path imgsz=(640, 640), # inference size h,w batch_size=1, # batch size dynamic=False, # dynamic batch size ): # PyTorch model im = torch.zeros((batch_size, 3, *imgsz)) # BCHW image model = attempt_load(weights, map_location=torch.device('cpu'), inplace=True, fuse=False) y = model(im) # inference model.info() # TensorFlow model im = tf.zeros((batch_size, *imgsz, 3)) # BHWC image tf_model = TFModel(cfg=model.yaml, model=model, nc=model.nc, imgsz=imgsz) y = tf_model.predict(im) # inference # Keras model im = keras.Input(shape=(*imgsz, 3), batch_size=None if dynamic else batch_size) keras_model = keras.Model(inputs=im, outputs=tf_model.predict(im)) keras_model.summary() LOGGER.info('PyTorch, TensorFlow and Keras models successfully verified.\nUse export.py for TF model export.') def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--weights', type=str, default=ROOT / 'yolov5s.pt', help='weights path') parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w') parser.add_argument('--batch-size', type=int, default=1, help='batch size') parser.add_argument('--dynamic', action='store_true', help='dynamic batch size') opt = parser.parse_args() opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand print_args(FILE.stem, opt) return opt def main(opt): run(**vars(opt)) if __name__ == "__main__": opt = parse_opt() main(opt) ================================================ FILE: models/yolo.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ YOLO-specific modules Usage: $ python path/to/models/yolo.py --cfg yolov5s.yaml """ import os import argparse import sys from copy import deepcopy from pathlib import Path FILE = Path(__file__).resolve() ROOT = FILE.parents[1] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH # ROOT = ROOT.relative_to(Path.cwd()) # relative from models.common import * from models.experimental import * from utils.autoanchor import check_anchor_order from utils.general import LOGGER, check_version, check_yaml, make_divisible, print_args from utils.plots import feature_visualization from utils.torch_utils import fuse_conv_and_bn, initialize_weights, model_info, scale_img, select_device, time_sync try: import thop # for FLOPs computation except ImportError: thop = None class Detect(nn.Module): stride = None # strides computed during build onnx_dynamic = False # ONNX export parameter def __init__(self, nc=80, anchors=(), ch=(), inplace=True): # detection layer super().__init__() self.nc = nc # number of classes self.no = nc + 5 # number of outputs per anchor self.nl = len(anchors) # number of detection layers self.na = len(anchors[0]) // 2 # number of anchors self.grid = [torch.zeros(1)] * self.nl # init grid self.anchor_grid = [torch.zeros(1)] * self.nl # init anchor grid self.register_buffer('anchors', torch.tensor(anchors).float().view(self.nl, -1, 2)) # shape(nl,na,2) self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv self.inplace = inplace # use in-place ops (e.g. slice assignment) def forward(self, x): z = [] # inference output for i in range(self.nl): x[i] = self.m[i](x[i]) # conv if os.getenv('RKNN_model_hack', '0') != '0': continue bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85) x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous() if not self.training: # inference if self.onnx_dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]: self.grid[i], self.anchor_grid[i] = self._make_grid(nx, ny, i) y = x[i].sigmoid() if self.inplace: y[..., 0:2] = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh y = torch.cat((xy, wh, y[..., 4:]), -1) z.append(y.view(bs, -1, self.no)) if os.getenv('RKNN_model_hack', '0') != '0': return x return x if self.training else (torch.cat(z, 1), x) def _make_grid(self, nx=20, ny=20, i=0): d = self.anchors[i].device if check_version(torch.__version__, '1.10.0'): # torch>=1.10.0 meshgrid workaround for torch>=0.7 compatibility yv, xv = torch.meshgrid([torch.arange(ny).to(d), torch.arange(nx).to(d)], indexing='ij') else: yv, xv = torch.meshgrid([torch.arange(ny).to(d), torch.arange(nx).to(d)]) grid = torch.stack((xv, yv), 2).expand((1, self.na, ny, nx, 2)).float() anchor_grid = (self.anchors[i].clone() * self.stride[i]) \ .view((1, self.na, 1, 1, 2)).expand((1, self.na, ny, nx, 2)).float() return grid, anchor_grid class Model(nn.Module): def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, anchors=None): # model, input channels, number of classes super().__init__() if isinstance(cfg, dict): self.yaml = cfg # model dict else: # is *.yaml import yaml # for torch hub self.yaml_file = Path(cfg).name with open(cfg, encoding='ascii', errors='ignore') as f: self.yaml = yaml.safe_load(f) # model dict # Define model ch = self.yaml['ch'] = self.yaml.get('ch', ch) # input channels if nc and nc != self.yaml['nc']: LOGGER.info(f"Overriding model.yaml nc={self.yaml['nc']} with nc={nc}") self.yaml['nc'] = nc # override yaml value if anchors: LOGGER.info(f'Overriding model.yaml anchors with anchors={anchors}') self.yaml['anchors'] = round(anchors) # override yaml value self.model, self.save = parse_model(deepcopy(self.yaml), ch=[ch]) # model, savelist self.names = [str(i) for i in range(self.yaml['nc'])] # default names self.inplace = self.yaml.get('inplace', True) # Build strides, anchors m = self.model[-1] # Detect() if isinstance(m, Detect): s = 256 # 2x min stride m.inplace = self.inplace m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))]) # forward m.anchors /= m.stride.view(-1, 1, 1) check_anchor_order(m) self.stride = m.stride self._initialize_biases() # only run once # Init weights, biases initialize_weights(self) self.info() LOGGER.info('') def forward(self, x, augment=False, profile=False, visualize=False): if augment: return self._forward_augment(x) # augmented inference, None return self._forward_once(x, profile, visualize) # single-scale inference, train def _forward_augment(self, x): img_size = x.shape[-2:] # height, width s = [1, 0.83, 0.67] # scales f = [None, 3, None] # flips (2-ud, 3-lr) y = [] # outputs for si, fi in zip(s, f): xi = scale_img(x.flip(fi) if fi else x, si, gs=int(self.stride.max())) yi = self._forward_once(xi)[0] # forward # cv2.imwrite(f'img_{si}.jpg', 255 * xi[0].cpu().numpy().transpose((1, 2, 0))[:, :, ::-1]) # save yi = self._descale_pred(yi, fi, si, img_size) y.append(yi) y = self._clip_augmented(y) # clip augmented tails return torch.cat(y, 1), None # augmented inference, train def _forward_once(self, x, profile=False, visualize=False): y, dt = [], [] # outputs for m in self.model: if m.f != -1: # if not from previous layer x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers if profile: self._profile_one_layer(m, x, dt) x = m(x) # run y.append(x if m.i in self.save else None) # save output if visualize: feature_visualization(x, m.type, m.i, save_dir=visualize) return x def _descale_pred(self, p, flips, scale, img_size): # de-scale predictions following augmented inference (inverse operation) if self.inplace: p[..., :4] /= scale # de-scale if flips == 2: p[..., 1] = img_size[0] - p[..., 1] # de-flip ud elif flips == 3: p[..., 0] = img_size[1] - p[..., 0] # de-flip lr else: x, y, wh = p[..., 0:1] / scale, p[..., 1:2] / scale, p[..., 2:4] / scale # de-scale if flips == 2: y = img_size[0] - y # de-flip ud elif flips == 3: x = img_size[1] - x # de-flip lr p = torch.cat((x, y, wh, p[..., 4:]), -1) return p def _clip_augmented(self, y): # Clip YOLOv5 augmented inference tails nl = self.model[-1].nl # number of detection layers (P3-P5) g = sum(4 ** x for x in range(nl)) # grid points e = 1 # exclude layer count i = (y[0].shape[1] // g) * sum(4 ** x for x in range(e)) # indices y[0] = y[0][:, :-i] # large i = (y[-1].shape[1] // g) * sum(4 ** (nl - 1 - x) for x in range(e)) # indices y[-1] = y[-1][:, i:] # small return y def _profile_one_layer(self, m, x, dt): c = isinstance(m, Detect) # is final layer, copy input as inplace fix o = thop.profile(m, inputs=(x.copy() if c else x,), verbose=False)[0] / 1E9 * 2 if thop else 0 # FLOPs t = time_sync() for _ in range(10): m(x.copy() if c else x) dt.append((time_sync() - t) * 100) if m == self.model[0]: LOGGER.info(f"{'time (ms)':>10s} {'GFLOPs':>10s} {'params':>10s} {'module'}") LOGGER.info(f'{dt[-1]:10.2f} {o:10.2f} {m.np:10.0f} {m.type}') if c: LOGGER.info(f"{sum(dt):10.2f} {'-':>10s} {'-':>10s} Total") def _initialize_biases(self, cf=None): # initialize biases into Detect(), cf is class frequency # https://arxiv.org/abs/1708.02002 section 3.3 # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1. m = self.model[-1] # Detect() module for mi, s in zip(m.m, m.stride): # from b = mi.bias.view(m.na, -1) # conv.bias(255) to (3,85) b.data[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image) b.data[:, 5:] += math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum()) # cls mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True) def _print_biases(self): m = self.model[-1] # Detect() module for mi in m.m: # from b = mi.bias.detach().view(m.na, -1).T # conv.bias(255) to (3,85) LOGGER.info( ('%6g Conv2d.bias:' + '%10.3g' * 6) % (mi.weight.shape[1], *b[:5].mean(1).tolist(), b[5:].mean())) # def _print_weights(self): # for m in self.model.modules(): # if type(m) is Bottleneck: # LOGGER.info('%10.3g' % (m.w.detach().sigmoid() * 2)) # shortcut weights def fuse(self): # fuse model Conv2d() + BatchNorm2d() layers LOGGER.info('Fusing layers... ') for m in self.model.modules(): if isinstance(m, (Conv, DWConv)) and hasattr(m, 'bn'): m.conv = fuse_conv_and_bn(m.conv, m.bn) # update conv delattr(m, 'bn') # remove batchnorm m.forward = m.forward_fuse # update forward self.info() return self def info(self, verbose=False, img_size=640): # print model information model_info(self, verbose, img_size) def _apply(self, fn): # Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers self = super()._apply(fn) m = self.model[-1] # Detect() if isinstance(m, Detect): m.stride = fn(m.stride) m.grid = list(map(fn, m.grid)) if isinstance(m.anchor_grid, list): m.anchor_grid = list(map(fn, m.anchor_grid)) return self def parse_model(d, ch): # model_dict, input_channels(3) LOGGER.info(f"\n{'':>3}{'from':>18}{'n':>3}{'params':>10} {'module':<40}{'arguments':<30}") anchors, nc, gd, gw = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple'] na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors # number of anchors no = na * (nc + 5) # number of outputs = anchors * (classes + 5) layers, save, c2 = [], [], ch[-1] # layers, savelist, ch out for i, (f, n, m, args) in enumerate(d['backbone'] + d['head']): # from, number, module, args m = eval(m) if isinstance(m, str) else m # eval strings for j, a in enumerate(args): try: args[j] = eval(a) if isinstance(a, str) else a # eval strings except NameError: pass n = n_ = max(round(n * gd), 1) if n > 1 else n # depth gain if m in [Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, BottleneckCSP, C3, C3TR, C3SPP, C3Ghost]: c1, c2 = ch[f], args[0] if c2 != no: # if not output c2 = make_divisible(c2 * gw, 8) args = [c1, c2, *args[1:]] if m in [BottleneckCSP, C3, C3TR, C3Ghost]: args.insert(2, n) # number of repeats n = 1 elif m is nn.BatchNorm2d: args = [ch[f]] elif m is Concat: c2 = sum(ch[x] for x in f) elif m is Detect: args.append([ch[x] for x in f]) if isinstance(args[1], int): # number of anchors args[1] = [list(range(args[1] * 2))] * len(f) elif m is Contract: c2 = ch[f] * args[0] ** 2 elif m is Expand: c2 = ch[f] // args[0] ** 2 else: c2 = ch[f] m_ = nn.Sequential(*(m(*args) for _ in range(n))) if n > 1 else m(*args) # module t = str(m)[8:-2].replace('__main__.', '') # module type np = sum(x.numel() for x in m_.parameters()) # number params m_.i, m_.f, m_.type, m_.np = i, f, t, np # attach index, 'from' index, type, number params LOGGER.info(f'{i:>3}{str(f):>18}{n_:>3}{np:10.0f} {t:<40}{str(args):<30}') # print save.extend(x % i for x in ([f] if isinstance(f, int) else f) if x != -1) # append to savelist layers.append(m_) if i == 0: ch = [] ch.append(c2) return nn.Sequential(*layers), sorted(save) if __name__ == '__main__': parser = argparse.ArgumentParser() parser.add_argument('--cfg', type=str, default='yolov5s.yaml', help='model.yaml') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--profile', action='store_true', help='profile model speed') parser.add_argument('--test', action='store_true', help='test all yolo*.yaml') opt = parser.parse_args() opt.cfg = check_yaml(opt.cfg) # check YAML print_args(FILE.stem, opt) device = select_device(opt.device) # Create model model = Model(opt.cfg).to(device) model.train() # Profile if opt.profile: img = torch.rand(8 if torch.cuda.is_available() else 1, 3, 640, 640).to(device) y = model(img, profile=True) # Test all models if opt.test: for cfg in Path(ROOT / 'models').rglob('yolo*.yaml'): try: _ = Model(cfg) except Exception as e: print(f'Error in {cfg}: {e}') # Tensorboard (not working https://github.com/ultralytics/yolov5/issues/2898) # from torch.utils.tensorboard import SummaryWriter # tb_writer = SummaryWriter('.') # LOGGER.info("Run 'tensorboard --logdir=models' to view tensorboard at http://localhost:6006/") # tb_writer.add_graph(torch.jit.trace(model, img, strict=False), []) # add model graph ================================================ FILE: models/yolov5l.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.0 # model depth multiple width_multiple: 1.0 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/yolov5m.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.67 # model depth multiple width_multiple: 0.75 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/yolov5n.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.33 # model depth multiple width_multiple: 0.25 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/yolov5s.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 0.33 # model depth multiple width_multiple: 0.50 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: models/yolov5x.yaml ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Parameters nc: 80 # number of classes depth_multiple: 1.33 # model depth multiple width_multiple: 1.25 # layer channel multiple anchors: - [10,13, 16,30, 33,23] # P3/8 - [30,61, 62,45, 59,119] # P4/16 - [116,90, 156,198, 373,326] # P5/32 # YOLOv5 v6.0 backbone backbone: # [from, number, module, args] [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 [-1, 3, C3, [128]], [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 [-1, 6, C3, [256]], [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 [-1, 9, C3, [512]], [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 [-1, 3, C3, [1024]], [-1, 1, SPPF, [1024, 5]], # 9 ] # YOLOv5 v6.0 head head: [[-1, 1, Conv, [512, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 6], 1, Concat, [1]], # cat backbone P4 [-1, 3, C3, [512, False]], # 13 [-1, 1, Conv, [256, 1, 1]], [-1, 1, nn.Upsample, [None, 2, 'nearest']], [[-1, 4], 1, Concat, [1]], # cat backbone P3 [-1, 3, C3, [256, False]], # 17 (P3/8-small) [-1, 1, Conv, [256, 3, 2]], [[-1, 14], 1, Concat, [1]], # cat head P4 [-1, 3, C3, [512, False]], # 20 (P4/16-medium) [-1, 1, Conv, [512, 3, 2]], [[-1, 10], 1, Concat, [1]], # cat head P5 [-1, 3, C3, [1024, False]], # 23 (P5/32-large) [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) ] ================================================ FILE: requirements.txt ================================================ # pip install -r requirements.txt # Base ---------------------------------------- matplotlib>=3.2.2 numpy>=1.18.5 opencv-python>=4.1.2 Pillow>=7.1.2 PyYAML>=5.3.1 requests>=2.23.0 scipy>=1.4.1 torch>=1.7.0 torchvision>=0.8.1 tqdm>=4.41.0 # Logging ------------------------------------- tensorboard>=2.4.1 # wandb # Plotting ------------------------------------ pandas>=1.1.4 seaborn>=0.11.0 # Export -------------------------------------- # coremltools>=4.1 # CoreML export # onnx>=1.9.0 # ONNX export # onnx-simplifier>=0.3.6 # ONNX simplifier # scikit-learn==0.19.2 # CoreML quantization # tensorflow>=2.4.1 # TFLite export # tensorflowjs>=3.9.0 # TF.js export # Extras -------------------------------------- # albumentations>=1.0.3 # Cython # for pycocotools https://github.com/cocodataset/cocoapi/issues/172 # pycocotools>=2.0 # COCO mAP # roboflow thop # FLOPs computation ================================================ FILE: setup.cfg ================================================ # Project-wide configuration file, can be used for package metadata and other toll configurations # Example usage: global configuration for PEP8 (via flake8) setting or default pytest arguments [metadata] license_file = LICENSE description-file = README.md [tool:pytest] norecursedirs = .git dist build addopts = --doctest-modules --durations=25 --color=yes [flake8] max-line-length = 120 exclude = .tox,*.egg,build,temp select = E,W,F doctests = True verbose = 2 # https://pep8.readthedocs.io/en/latest/intro.html#error-codes format = pylint # see: https://www.flake8rules.com/ ignore = E731 # Do not assign a lambda expression, use a def F405 E402 F841 E741 F821 E722 F401 W504 E127 W504 E231 E501 F403 E302 F541 [isort] # https://pycqa.github.io/isort/docs/configuration/options.html line_length = 120 multi_line_output = 0 ================================================ FILE: train.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Train a YOLOv5 model on a custom dataset Usage: $ python path/to/train.py --data coco128.yaml --weights yolov5s.pt --img 640 """ import argparse import math import os import random import sys import time from copy import deepcopy from datetime import datetime from pathlib import Path import numpy as np import torch import torch.distributed as dist import torch.nn as nn import yaml from torch.cuda import amp from torch.nn.parallel import DistributedDataParallel as DDP from torch.optim import SGD, Adam, lr_scheduler from tqdm import tqdm FILE = Path(__file__).resolve() ROOT = FILE.parents[0] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative import val # for end-of-epoch mAP from models.experimental import attempt_load from models.yolo import Model from utils.autoanchor import check_anchors from utils.autobatch import check_train_batch_size from utils.callbacks import Callbacks from utils.datasets import create_dataloader from utils.downloads import attempt_download from utils.general import (LOGGER, check_dataset, check_file, check_git_status, check_img_size, check_requirements, check_suffix, check_yaml, colorstr, get_latest_run, increment_path, init_seeds, intersect_dicts, labels_to_class_weights, labels_to_image_weights, methods, one_cycle, print_args, print_mutation, strip_optimizer) from utils.loggers import Loggers from utils.loggers.wandb.wandb_utils import check_wandb_resume from utils.loss import ComputeLoss from utils.metrics import fitness from utils.plots import plot_evolve, plot_labels from utils.torch_utils import EarlyStopping, ModelEMA, de_parallel, select_device, torch_distributed_zero_first LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html RANK = int(os.getenv('RANK', -1)) WORLD_SIZE = int(os.getenv('WORLD_SIZE', 1)) def train(hyp, # path/to/hyp.yaml or hyp dictionary opt, device, callbacks ): save_dir, epochs, batch_size, weights, single_cls, evolve, data, cfg, resume, noval, nosave, workers, freeze, = \ Path(opt.save_dir), opt.epochs, opt.batch_size, opt.weights, opt.single_cls, opt.evolve, opt.data, opt.cfg, \ opt.resume, opt.noval, opt.nosave, opt.workers, opt.freeze # Directories w = save_dir / 'weights' # weights dir (w.parent if evolve else w).mkdir(parents=True, exist_ok=True) # make dir last, best = w / 'last.pt', w / 'best.pt' # Hyperparameters if isinstance(hyp, str): with open(hyp, errors='ignore') as f: hyp = yaml.safe_load(f) # load hyps dict LOGGER.info(colorstr('hyperparameters: ') + ', '.join(f'{k}={v}' for k, v in hyp.items())) # Save run settings if not evolve: with open(save_dir / 'hyp.yaml', 'w') as f: yaml.safe_dump(hyp, f, sort_keys=False) with open(save_dir / 'opt.yaml', 'w') as f: yaml.safe_dump(vars(opt), f, sort_keys=False) # Loggers data_dict = None if RANK in [-1, 0]: loggers = Loggers(save_dir, weights, opt, hyp, LOGGER) # loggers instance if loggers.wandb: data_dict = loggers.wandb.data_dict if resume: weights, epochs, hyp = opt.weights, opt.epochs, opt.hyp # Register actions for k in methods(loggers): callbacks.register_action(k, callback=getattr(loggers, k)) # Config plots = not evolve # create plots cuda = device.type != 'cpu' init_seeds(1 + RANK) with torch_distributed_zero_first(LOCAL_RANK): data_dict = data_dict or check_dataset(data) # check if None train_path, val_path = data_dict['train'], data_dict['val'] nc = 1 if single_cls else int(data_dict['nc']) # number of classes names = ['item'] if single_cls and len(data_dict['names']) != 1 else data_dict['names'] # class names assert len(names) == nc, f'{len(names)} names found for nc={nc} dataset in {data}' # check is_coco = isinstance(val_path, str) and val_path.endswith('coco/val2017.txt') # COCO dataset # Model check_suffix(weights, '.pt') # check weights pretrained = weights.endswith('.pt') if pretrained: with torch_distributed_zero_first(LOCAL_RANK): weights = attempt_download(weights) # download if not found locally ckpt = torch.load(weights, map_location=device) # load checkpoint model = Model(cfg or ckpt['model'].yaml, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) # create exclude = ['anchor'] if (cfg or hyp.get('anchors')) and not resume else [] # exclude keys csd = ckpt['model'].float().state_dict() # checkpoint state_dict as FP32 csd = intersect_dicts(csd, model.state_dict(), exclude=exclude) # intersect model.load_state_dict(csd, strict=False) # load LOGGER.info(f'Transferred {len(csd)}/{len(model.state_dict())} items from {weights}') # report else: model = Model(cfg, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) # create # Freeze freeze = [f'model.{x}.' for x in range(freeze)] # layers to freeze for k, v in model.named_parameters(): v.requires_grad = True # train all layers if any(x in k for x in freeze): LOGGER.info(f'freezing {k}') v.requires_grad = False # Image size gs = max(int(model.stride.max()), 32) # grid size (max stride) imgsz = check_img_size(opt.imgsz, gs, floor=gs * 2) # verify imgsz is gs-multiple # Batch size if RANK == -1 and batch_size == -1: # single-GPU only, estimate best batch size batch_size = check_train_batch_size(model, imgsz) # Optimizer nbs = 64 # nominal batch size accumulate = max(round(nbs / batch_size), 1) # accumulate loss before optimizing hyp['weight_decay'] *= batch_size * accumulate / nbs # scale weight_decay LOGGER.info(f"Scaled weight_decay = {hyp['weight_decay']}") g0, g1, g2 = [], [], [] # optimizer parameter groups for v in model.modules(): if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter): # bias g2.append(v.bias) if isinstance(v, nn.BatchNorm2d): # weight (no decay) g0.append(v.weight) elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter): # weight (with decay) g1.append(v.weight) if opt.adam: optimizer = Adam(g0, lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum else: optimizer = SGD(g0, lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True) optimizer.add_param_group({'params': g1, 'weight_decay': hyp['weight_decay']}) # add g1 with weight_decay optimizer.add_param_group({'params': g2}) # add g2 (biases) LOGGER.info(f"{colorstr('optimizer:')} {type(optimizer).__name__} with parameter groups " f"{len(g0)} weight, {len(g1)} weight (no decay), {len(g2)} bias") del g0, g1, g2 # Scheduler if opt.linear_lr: lf = lambda x: (1 - x / (epochs - 1)) * (1.0 - hyp['lrf']) + hyp['lrf'] # linear else: lf = one_cycle(1, hyp['lrf'], epochs) # cosine 1->hyp['lrf'] scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lf) # plot_lr_scheduler(optimizer, scheduler, epochs) # EMA ema = ModelEMA(model) if RANK in [-1, 0] else None # Resume start_epoch, best_fitness = 0, 0.0 if pretrained: # Optimizer if ckpt['optimizer'] is not None: optimizer.load_state_dict(ckpt['optimizer']) best_fitness = ckpt['best_fitness'] # EMA if ema and ckpt.get('ema'): ema.ema.load_state_dict(ckpt['ema'].float().state_dict()) ema.updates = ckpt['updates'] # Epochs start_epoch = ckpt['epoch'] + 1 if resume: assert start_epoch > 0, f'{weights} training to {epochs} epochs is finished, nothing to resume.' if epochs < start_epoch: LOGGER.info(f"{weights} has been trained for {ckpt['epoch']} epochs. Fine-tuning for {epochs} more epochs.") epochs += ckpt['epoch'] # finetune additional epochs del ckpt, csd # DP mode if cuda and RANK == -1 and torch.cuda.device_count() > 1: LOGGER.warning('WARNING: DP not recommended, use torch.distributed.run for best DDP Multi-GPU results.\n' 'See Multi-GPU Tutorial at https://github.com/ultralytics/yolov5/issues/475 to get started.') model = torch.nn.DataParallel(model) # SyncBatchNorm if opt.sync_bn and cuda and RANK != -1: model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model).to(device) LOGGER.info('Using SyncBatchNorm()') # Trainloader train_loader, dataset = create_dataloader(train_path, imgsz, batch_size // WORLD_SIZE, gs, single_cls, hyp=hyp, augment=True, cache=opt.cache, rect=opt.rect, rank=LOCAL_RANK, workers=workers, image_weights=opt.image_weights, quad=opt.quad, prefix=colorstr('train: '), shuffle=True) mlc = int(np.concatenate(dataset.labels, 0)[:, 0].max()) # max label class nb = len(train_loader) # number of batches assert mlc < nc, f'Label class {mlc} exceeds nc={nc} in {data}. Possible class labels are 0-{nc - 1}' # Process 0 if RANK in [-1, 0]: val_loader = create_dataloader(val_path, imgsz, batch_size // WORLD_SIZE * 2, gs, single_cls, hyp=hyp, cache=None if noval else opt.cache, rect=True, rank=-1, workers=workers, pad=0.5, prefix=colorstr('val: '))[0] if not resume: labels = np.concatenate(dataset.labels, 0) # c = torch.tensor(labels[:, 0]) # classes # cf = torch.bincount(c.long(), minlength=nc) + 1. # frequency # model._initialize_biases(cf.to(device)) if plots: plot_labels(labels, names, save_dir) # Anchors if not opt.noautoanchor: check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz) model.half().float() # pre-reduce anchor precision callbacks.run('on_pretrain_routine_end') # DDP mode if cuda and RANK != -1: model = DDP(model, device_ids=[LOCAL_RANK], output_device=LOCAL_RANK) # Model attributes nl = de_parallel(model).model[-1].nl # number of detection layers (to scale hyps) hyp['box'] *= 3 / nl # scale to layers hyp['cls'] *= nc / 80 * 3 / nl # scale to classes and layers hyp['obj'] *= (imgsz / 640) ** 2 * 3 / nl # scale to image size and layers hyp['label_smoothing'] = opt.label_smoothing model.nc = nc # attach number of classes to model model.hyp = hyp # attach hyperparameters to model model.class_weights = labels_to_class_weights(dataset.labels, nc).to(device) * nc # attach class weights model.names = names # Start training t0 = time.time() nw = max(round(hyp['warmup_epochs'] * nb), 1000) # number of warmup iterations, max(3 epochs, 1k iterations) # nw = min(nw, (epochs - start_epoch) / 2 * nb) # limit warmup to < 1/2 of training last_opt_step = -1 maps = np.zeros(nc) # mAP per class results = (0, 0, 0, 0, 0, 0, 0) # P, R, mAP@.5, mAP@.5-.95, val_loss(box, obj, cls) scheduler.last_epoch = start_epoch - 1 # do not move scaler = amp.GradScaler(enabled=cuda) stopper = EarlyStopping(patience=opt.patience) compute_loss = ComputeLoss(model) # init loss class LOGGER.info(f'Image sizes {imgsz} train, {imgsz} val\n' f'Using {train_loader.num_workers * WORLD_SIZE} dataloader workers\n' f"Logging results to {colorstr('bold', save_dir)}\n" f'Starting training for {epochs} epochs...') for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------ model.train() # Update image weights (optional, single-GPU only) if opt.image_weights: cw = model.class_weights.cpu().numpy() * (1 - maps) ** 2 / nc # class weights iw = labels_to_image_weights(dataset.labels, nc=nc, class_weights=cw) # image weights dataset.indices = random.choices(range(dataset.n), weights=iw, k=dataset.n) # rand weighted idx # Update mosaic border (optional) # b = int(random.uniform(0.25 * imgsz, 0.75 * imgsz + gs) // gs * gs) # dataset.mosaic_border = [b - imgsz, -b] # height, width borders mloss = torch.zeros(3, device=device) # mean losses if RANK != -1: train_loader.sampler.set_epoch(epoch) pbar = enumerate(train_loader) LOGGER.info(('\n' + '%10s' * 7) % ('Epoch', 'gpu_mem', 'box', 'obj', 'cls', 'labels', 'img_size')) if RANK in [-1, 0]: pbar = tqdm(pbar, total=nb, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar optimizer.zero_grad() for i, (imgs, targets, paths, _) in pbar: # batch ------------------------------------------------------------- ni = i + nb * epoch # number integrated batches (since train start) imgs = imgs.to(device, non_blocking=True).float() / 255 # uint8 to float32, 0-255 to 0.0-1.0 # Warmup if ni <= nw: xi = [0, nw] # x interp # compute_loss.gr = np.interp(ni, xi, [0.0, 1.0]) # iou loss ratio (obj_loss = 1.0 or iou) accumulate = max(1, np.interp(ni, xi, [1, nbs / batch_size]).round()) for j, x in enumerate(optimizer.param_groups): # bias lr falls from 0.1 to lr0, all other lrs rise from 0.0 to lr0 x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 2 else 0.0, x['initial_lr'] * lf(epoch)]) if 'momentum' in x: x['momentum'] = np.interp(ni, xi, [hyp['warmup_momentum'], hyp['momentum']]) # Multi-scale if opt.multi_scale: sz = random.randrange(imgsz * 0.5, imgsz * 1.5 + gs) // gs * gs # size sf = sz / max(imgs.shape[2:]) # scale factor if sf != 1: ns = [math.ceil(x * sf / gs) * gs for x in imgs.shape[2:]] # new shape (stretched to gs-multiple) imgs = nn.functional.interpolate(imgs, size=ns, mode='bilinear', align_corners=False) # Forward with amp.autocast(enabled=cuda): pred = model(imgs) # forward loss, loss_items = compute_loss(pred, targets.to(device)) # loss scaled by batch_size if RANK != -1: loss *= WORLD_SIZE # gradient averaged between devices in DDP mode if opt.quad: loss *= 4. # Backward scaler.scale(loss).backward() # Optimize if ni - last_opt_step >= accumulate: scaler.step(optimizer) # optimizer.step scaler.update() optimizer.zero_grad() if ema: ema.update(model) last_opt_step = ni # Log if RANK in [-1, 0]: mloss = (mloss * i + loss_items) / (i + 1) # update mean losses mem = f'{torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0:.3g}G' # (GB) pbar.set_description(('%10s' * 2 + '%10.4g' * 5) % ( f'{epoch}/{epochs - 1}', mem, *mloss, targets.shape[0], imgs.shape[-1])) callbacks.run('on_train_batch_end', ni, model, imgs, targets, paths, plots, opt.sync_bn) # end batch ------------------------------------------------------------------------------------------------ # Scheduler lr = [x['lr'] for x in optimizer.param_groups] # for loggers scheduler.step() if RANK in [-1, 0]: # mAP callbacks.run('on_train_epoch_end', epoch=epoch) ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'names', 'stride', 'class_weights']) final_epoch = (epoch + 1 == epochs) or stopper.possible_stop if not noval or final_epoch: # Calculate mAP results, maps, _ = val.run(data_dict, batch_size=batch_size // WORLD_SIZE * 2, imgsz=imgsz, model=ema.ema, single_cls=single_cls, dataloader=val_loader, save_dir=save_dir, plots=False, callbacks=callbacks, compute_loss=compute_loss) # Update best mAP fi = fitness(np.array(results).reshape(1, -1)) # weighted combination of [P, R, mAP@.5, mAP@.5-.95] if fi > best_fitness: best_fitness = fi log_vals = list(mloss) + list(results) + lr callbacks.run('on_fit_epoch_end', log_vals, epoch, best_fitness, fi) # Save model if (not nosave) or (final_epoch and not evolve): # if save ckpt = {'epoch': epoch, 'best_fitness': best_fitness, 'model': deepcopy(de_parallel(model)).half(), 'ema': deepcopy(ema.ema).half(), 'updates': ema.updates, 'optimizer': optimizer.state_dict(), 'wandb_id': loggers.wandb.wandb_run.id if loggers.wandb else None, 'date': datetime.now().isoformat()} # Save last, best and delete torch.save(ckpt, last) if best_fitness == fi: torch.save(ckpt, best) if (epoch > 0) and (opt.save_period > 0) and (epoch % opt.save_period == 0): torch.save(ckpt, w / f'epoch{epoch}.pt') del ckpt callbacks.run('on_model_save', last, epoch, final_epoch, best_fitness, fi) # Stop Single-GPU if RANK == -1 and stopper(epoch=epoch, fitness=fi): break # Stop DDP TODO: known issues shttps://github.com/ultralytics/yolov5/pull/4576 # stop = stopper(epoch=epoch, fitness=fi) # if RANK == 0: # dist.broadcast_object_list([stop], 0) # broadcast 'stop' to all ranks # Stop DPP # with torch_distributed_zero_first(RANK): # if stop: # break # must break all DDP ranks # end epoch ---------------------------------------------------------------------------------------------------- # end training ----------------------------------------------------------------------------------------------------- if RANK in [-1, 0]: LOGGER.info(f'\n{epoch - start_epoch + 1} epochs completed in {(time.time() - t0) / 3600:.3f} hours.') for f in last, best: if f.exists(): strip_optimizer(f) # strip optimizers if f is best: LOGGER.info(f'\nValidating {f}...') results, _, _ = val.run(data_dict, batch_size=batch_size // WORLD_SIZE * 2, imgsz=imgsz, model=attempt_load(f, device).half(), iou_thres=0.65 if is_coco else 0.60, # best pycocotools results at 0.65 single_cls=single_cls, dataloader=val_loader, save_dir=save_dir, save_json=is_coco, verbose=True, plots=True, callbacks=callbacks, compute_loss=compute_loss) # val best model with plots if is_coco: callbacks.run('on_fit_epoch_end', list(mloss) + list(results) + lr, epoch, best_fitness, fi) callbacks.run('on_train_end', last, best, plots, epoch, results) LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") torch.cuda.empty_cache() return results def parse_opt(known=False): parser = argparse.ArgumentParser() parser.add_argument('--weights', type=str, default=ROOT / 'yolov5s.pt', help='initial weights path') parser.add_argument('--cfg', type=str, default='', help='model.yaml path') parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') parser.add_argument('--hyp', type=str, default=ROOT / 'data/hyps/hyp.scratch.yaml', help='hyperparameters path') parser.add_argument('--epochs', type=int, default=300) parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs, -1 for autobatch') parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='train, val image size (pixels)') parser.add_argument('--rect', action='store_true', help='rectangular training') parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training') parser.add_argument('--nosave', action='store_true', help='only save final checkpoint') parser.add_argument('--noval', action='store_true', help='only validate final epoch') parser.add_argument('--noautoanchor', action='store_true', help='disable autoanchor check') parser.add_argument('--evolve', type=int, nargs='?', const=300, help='evolve hyperparameters for x generations') parser.add_argument('--bucket', type=str, default='', help='gsutil bucket') parser.add_argument('--cache', type=str, nargs='?', const='ram', help='--cache images in "ram" (default) or "disk"') parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--multi-scale', action='store_true', help='vary img-size +/- 50%%') parser.add_argument('--single-cls', action='store_true', help='train multi-class data as single-class') parser.add_argument('--adam', action='store_true', help='use torch.optim.Adam() optimizer') parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode') parser.add_argument('--workers', type=int, default=2, help='max dataloader workers (per RANK in DDP mode)') parser.add_argument('--project', default=ROOT / 'runs/train', help='save to project/name') parser.add_argument('--name', default='exp', help='save to project/name') parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') parser.add_argument('--quad', action='store_true', help='quad dataloader') parser.add_argument('--linear-lr', action='store_true', help='linear LR') parser.add_argument('--label-smoothing', type=float, default=0.0, help='Label smoothing epsilon') parser.add_argument('--patience', type=int, default=100, help='EarlyStopping patience (epochs without improvement)') parser.add_argument('--freeze', type=int, default=0, help='Number of layers to freeze. backbone=10, all=24') parser.add_argument('--save-period', type=int, default=-1, help='Save checkpoint every x epochs (disabled if < 1)') parser.add_argument('--local_rank', type=int, default=-1, help='DDP parameter, do not modify') # Weights & Biases arguments parser.add_argument('--entity', default=None, help='W&B: Entity') parser.add_argument('--upload_dataset', nargs='?', const=True, default=False, help='W&B: Upload data, "val" option') parser.add_argument('--bbox_interval', type=int, default=-1, help='W&B: Set bounding-box image logging interval') parser.add_argument('--artifact_alias', type=str, default='latest', help='W&B: Version of dataset artifact to use') opt = parser.parse_known_args()[0] if known else parser.parse_args() return opt def main(opt, callbacks=Callbacks()): # Checks if RANK in [-1, 0]: print_args(FILE.stem, opt) check_git_status() check_requirements(exclude=['thop']) # Resume if opt.resume and not check_wandb_resume(opt) and not opt.evolve: # resume an interrupted run ckpt = opt.resume if isinstance(opt.resume, str) else get_latest_run() # specified or most recent path assert os.path.isfile(ckpt), 'ERROR: --resume checkpoint does not exist' with open(Path(ckpt).parent.parent / 'opt.yaml', errors='ignore') as f: opt = argparse.Namespace(**yaml.safe_load(f)) # replace opt.cfg, opt.weights, opt.resume = '', ckpt, True # reinstate LOGGER.info(f'Resuming training from {ckpt}') else: opt.data, opt.cfg, opt.hyp, opt.weights, opt.project = \ check_file(opt.data), check_yaml(opt.cfg), check_yaml(opt.hyp), str(opt.weights), str(opt.project) # checks assert len(opt.cfg) or len(opt.weights), 'either --cfg or --weights must be specified' if opt.evolve: opt.project = str(ROOT / 'runs/evolve') opt.exist_ok, opt.resume = opt.resume, False # pass resume to exist_ok and disable resume opt.save_dir = str(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok)) # DDP mode device = select_device(opt.device, batch_size=opt.batch_size) if LOCAL_RANK != -1: assert torch.cuda.device_count() > LOCAL_RANK, 'insufficient CUDA devices for DDP command' assert opt.batch_size % WORLD_SIZE == 0, '--batch-size must be multiple of CUDA device count' assert not opt.image_weights, '--image-weights argument is not compatible with DDP training' assert not opt.evolve, '--evolve argument is not compatible with DDP training' torch.cuda.set_device(LOCAL_RANK) device = torch.device('cuda', LOCAL_RANK) dist.init_process_group(backend="nccl" if dist.is_nccl_available() else "gloo") # Train if not opt.evolve: train(opt.hyp, opt, device, callbacks) if WORLD_SIZE > 1 and RANK == 0: LOGGER.info('Destroying process group... ') dist.destroy_process_group() # Evolve hyperparameters (optional) else: # Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit) meta = {'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3) 'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf) 'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1 'weight_decay': (1, 0.0, 0.001), # optimizer weight decay 'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok) 'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum 'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr 'box': (1, 0.02, 0.2), # box loss gain 'cls': (1, 0.2, 4.0), # cls loss gain 'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight 'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels) 'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight 'iou_t': (0, 0.1, 0.7), # IoU training threshold 'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold 'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore) 'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5) 'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction) 'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction) 'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction) 'degrees': (1, 0.0, 45.0), # image rotation (+/- deg) 'translate': (1, 0.0, 0.9), # image translation (+/- fraction) 'scale': (1, 0.0, 0.9), # image scale (+/- gain) 'shear': (1, 0.0, 10.0), # image shear (+/- deg) 'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001 'flipud': (1, 0.0, 1.0), # image flip up-down (probability) 'fliplr': (0, 0.0, 1.0), # image flip left-right (probability) 'mosaic': (1, 0.0, 1.0), # image mixup (probability) 'mixup': (1, 0.0, 1.0), # image mixup (probability) 'copy_paste': (1, 0.0, 1.0)} # segment copy-paste (probability) with open(opt.hyp, errors='ignore') as f: hyp = yaml.safe_load(f) # load hyps dict if 'anchors' not in hyp: # anchors commented in hyp.yaml hyp['anchors'] = 3 opt.noval, opt.nosave, save_dir = True, True, Path(opt.save_dir) # only val/save final epoch # ei = [isinstance(x, (int, float)) for x in hyp.values()] # evolvable indices evolve_yaml, evolve_csv = save_dir / 'hyp_evolve.yaml', save_dir / 'evolve.csv' if opt.bucket: os.system(f'gsutil cp gs://{opt.bucket}/evolve.csv {save_dir}') # download evolve.csv if exists for _ in range(opt.evolve): # generations to evolve if evolve_csv.exists(): # if evolve.csv exists: select best hyps and mutate # Select parent(s) parent = 'single' # parent selection method: 'single' or 'weighted' x = np.loadtxt(evolve_csv, ndmin=2, delimiter=',', skiprows=1) n = min(5, len(x)) # number of previous results to consider x = x[np.argsort(-fitness(x))][:n] # top n mutations w = fitness(x) - fitness(x).min() + 1E-6 # weights (sum > 0) if parent == 'single' or len(x) == 1: # x = x[random.randint(0, n - 1)] # random selection x = x[random.choices(range(n), weights=w)[0]] # weighted selection elif parent == 'weighted': x = (x * w.reshape(n, 1)).sum(0) / w.sum() # weighted combination # Mutate mp, s = 0.8, 0.2 # mutation probability, sigma npr = np.random npr.seed(int(time.time())) g = np.array([meta[k][0] for k in hyp.keys()]) # gains 0-1 ng = len(meta) v = np.ones(ng) while all(v == 1): # mutate until a change occurs (prevent duplicates) v = (g * (npr.random(ng) < mp) * npr.randn(ng) * npr.random() * s + 1).clip(0.3, 3.0) for i, k in enumerate(hyp.keys()): # plt.hist(v.ravel(), 300) hyp[k] = float(x[i + 7] * v[i]) # mutate # Constrain to limits for k, v in meta.items(): hyp[k] = max(hyp[k], v[1]) # lower limit hyp[k] = min(hyp[k], v[2]) # upper limit hyp[k] = round(hyp[k], 5) # significant digits # Train mutation results = train(hyp.copy(), opt, device, callbacks) # Write mutation results print_mutation(results, hyp.copy(), save_dir, opt.bucket) # Plot results plot_evolve(evolve_csv) LOGGER.info(f'Hyperparameter evolution finished\n' f"Results saved to {colorstr('bold', save_dir)}\n" f'Use best hyperparameters example: $ python train.py --hyp {evolve_yaml}') def run(**kwargs): # Usage: import train; train.run(data='coco128.yaml', imgsz=320, weights='yolov5m.pt') opt = parse_opt(True) for k, v in kwargs.items(): setattr(opt, k, v) main(opt) if __name__ == "__main__": opt = parse_opt() main(opt) ================================================ FILE: tutorial.ipynb ================================================ { "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "name": "YOLOv5 Tutorial", "provenance": [], "collapsed_sections": [], "include_colab_link": true }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "accelerator": "GPU", "widgets": { "application/vnd.jupyter.widget-state+json": { "eb95db7cae194218b3fcefb439b6352f": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_view_name": "HBoxView", "_dom_classes": [], "_model_name": "HBoxModel", "_view_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_view_count": null, "_view_module_version": "1.5.0", "box_style": "", "layout": "IPY_MODEL_769ecde6f2e64bacb596ce972f8d3d2d", "_model_module": "@jupyter-widgets/controls", "children": [ "IPY_MODEL_384a001876054c93b0af45cd1e960bfe", "IPY_MODEL_dded0aeae74440f7ba2ffa0beb8dd612", "IPY_MODEL_5296d28be75740b2892ae421bbec3657" ] } }, "769ecde6f2e64bacb596ce972f8d3d2d": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_view_name": "LayoutView", "grid_template_rows": null, "right": null, "justify_content": null, "_view_module": "@jupyter-widgets/base", "overflow": null, "_model_module_version": "1.2.0", "_view_count": null, "flex_flow": null, "width": null, "min_width": null, "border": null, "align_items": null, "bottom": null, "_model_module": "@jupyter-widgets/base", "top": null, "grid_column": null, "overflow_y": null, "overflow_x": null, "grid_auto_flow": null, "grid_area": null, "grid_template_columns": null, "flex": null, "_model_name": "LayoutModel", "justify_items": null, "grid_row": null, "max_height": null, "align_content": null, "visibility": null, "align_self": null, "height": null, "min_height": null, "padding": null, "grid_auto_rows": null, "grid_gap": null, "max_width": null, "order": null, "_view_module_version": "1.2.0", "grid_template_areas": null, "object_position": null, "object_fit": null, "grid_auto_columns": null, "margin": null, "display": null, "left": null } }, "384a001876054c93b0af45cd1e960bfe": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_view_name": "HTMLView", "style": "IPY_MODEL_9f09facb2a6c4a7096810d327c8b551c", "_dom_classes": [], "description": "", "_model_name": "HTMLModel", "placeholder": "​", "_view_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "value": "100%", "_view_count": null, "_view_module_version": "1.5.0", "description_tooltip": null, "_model_module": "@jupyter-widgets/controls", "layout": "IPY_MODEL_25621cff5d16448cb7260e839fd0f543" } }, "dded0aeae74440f7ba2ffa0beb8dd612": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_view_name": "ProgressView", "style": "IPY_MODEL_0ce7164fc0c74bb9a2b5c7037375a727", "_dom_classes": [], "description": "", "_model_name": "FloatProgressModel", "bar_style": "success", "max": 818322941, "_view_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "value": 818322941, "_view_count": null, "_view_module_version": "1.5.0", "orientation": "horizontal", "min": 0, "description_tooltip": null, "_model_module": "@jupyter-widgets/controls", "layout": "IPY_MODEL_c4c4593c10904cb5b8a5724d60c7e181" } }, "5296d28be75740b2892ae421bbec3657": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_view_name": "HTMLView", "style": "IPY_MODEL_473371611126476c88d5d42ec7031ed6", "_dom_classes": [], "description": "", "_model_name": "HTMLModel", "placeholder": "​", "_view_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "value": " 780M/780M [00:11<00:00, 91.9MB/s]", "_view_count": null, "_view_module_version": "1.5.0", "description_tooltip": null, "_model_module": "@jupyter-widgets/controls", "layout": "IPY_MODEL_65efdfd0d26c46e79c8c5ff3b77126cc" } }, "9f09facb2a6c4a7096810d327c8b551c": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_view_name": "StyleView", "_model_name": "DescriptionStyleModel", "description_width": "", "_view_module": "@jupyter-widgets/base", "_model_module_version": "1.5.0", "_view_count": null, "_view_module_version": "1.2.0", "_model_module": "@jupyter-widgets/controls" } }, "25621cff5d16448cb7260e839fd0f543": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_view_name": "LayoutView", "grid_template_rows": null, "right": null, "justify_content": null, "_view_module": "@jupyter-widgets/base", "overflow": null, "_model_module_version": "1.2.0", "_view_count": null, "flex_flow": null, "width": null, "min_width": null, "border": null, "align_items": null, "bottom": null, "_model_module": "@jupyter-widgets/base", "top": null, "grid_column": null, "overflow_y": null, "overflow_x": null, "grid_auto_flow": null, "grid_area": null, "grid_template_columns": null, "flex": null, "_model_name": "LayoutModel", "justify_items": null, "grid_row": null, "max_height": null, "align_content": null, "visibility": null, "align_self": null, "height": null, "min_height": null, "padding": null, "grid_auto_rows": null, "grid_gap": null, "max_width": null, "order": null, "_view_module_version": "1.2.0", "grid_template_areas": null, "object_position": null, "object_fit": null, "grid_auto_columns": null, "margin": null, "display": null, "left": null } }, "0ce7164fc0c74bb9a2b5c7037375a727": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_view_name": "StyleView", "_model_name": "ProgressStyleModel", "description_width": "", "_view_module": "@jupyter-widgets/base", "_model_module_version": "1.5.0", "_view_count": null, "_view_module_version": "1.2.0", "bar_color": null, "_model_module": "@jupyter-widgets/controls" } }, "c4c4593c10904cb5b8a5724d60c7e181": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_view_name": "LayoutView", "grid_template_rows": null, "right": null, "justify_content": null, "_view_module": "@jupyter-widgets/base", "overflow": null, "_model_module_version": "1.2.0", "_view_count": null, "flex_flow": null, "width": null, "min_width": null, "border": null, "align_items": null, "bottom": null, "_model_module": "@jupyter-widgets/base", "top": null, "grid_column": null, "overflow_y": null, "overflow_x": null, "grid_auto_flow": null, "grid_area": null, "grid_template_columns": null, "flex": null, "_model_name": "LayoutModel", "justify_items": null, "grid_row": null, "max_height": null, "align_content": null, "visibility": null, "align_self": null, "height": null, "min_height": null, "padding": null, "grid_auto_rows": null, "grid_gap": null, "max_width": null, "order": null, "_view_module_version": "1.2.0", "grid_template_areas": null, "object_position": null, "object_fit": null, "grid_auto_columns": null, "margin": null, "display": null, "left": null } }, "473371611126476c88d5d42ec7031ed6": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_view_name": "StyleView", "_model_name": "DescriptionStyleModel", "description_width": "", "_view_module": "@jupyter-widgets/base", "_model_module_version": "1.5.0", "_view_count": null, "_view_module_version": "1.2.0", "_model_module": "@jupyter-widgets/controls" } }, "65efdfd0d26c46e79c8c5ff3b77126cc": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_view_name": "LayoutView", "grid_template_rows": null, "right": null, "justify_content": null, "_view_module": "@jupyter-widgets/base", "overflow": null, "_model_module_version": "1.2.0", "_view_count": null, "flex_flow": null, "width": null, "min_width": null, "border": null, "align_items": null, "bottom": null, "_model_module": "@jupyter-widgets/base", "top": null, "grid_column": null, "overflow_y": null, "overflow_x": null, "grid_auto_flow": null, "grid_area": null, "grid_template_columns": null, "flex": null, "_model_name": "LayoutModel", "justify_items": null, "grid_row": null, "max_height": null, "align_content": null, "visibility": null, "align_self": null, "height": null, "min_height": null, "padding": null, "grid_auto_rows": null, "grid_gap": null, "max_width": null, "order": null, "_view_module_version": "1.2.0", "grid_template_areas": null, "object_position": null, "object_fit": null, "grid_auto_columns": null, "margin": null, "display": null, "left": null } } } } }, "cells": [ { "cell_type": "markdown", "metadata": { "id": "view-in-github", "colab_type": "text" }, "source": [ "\"Open" ] }, { "cell_type": "markdown", "metadata": { "id": "t6MPjfT5NrKQ" }, "source": [ "\n", "\n", "\n", "This is the **official YOLOv5 🚀 notebook** by **Ultralytics**, and is freely available for redistribution under the [GPL-3.0 license](https://choosealicense.com/licenses/gpl-3.0/). \n", "For more information please visit https://github.com/ultralytics/yolov5 and https://ultralytics.com. Thank you!" ] }, { "cell_type": "markdown", "metadata": { "id": "7mGmQbAO5pQb" }, "source": [ "# Setup\n", "\n", "Clone repo, install dependencies and check PyTorch and GPU." ] }, { "cell_type": "code", "metadata": { "id": "wbvMlHd_QwMG", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "3809e5a9-dd41-4577-fe62-5531abf7cca2" }, "source": [ "!git clone https://github.com/ultralytics/yolov5 # clone\n", "%cd yolov5\n", "%pip install -qr requirements.txt # install\n", "\n", "import torch\n", "from yolov5 import utils\n", "display = utils.notebook_init() # checks" ], "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "YOLOv5 🚀 v6.0-48-g84a8099 torch 1.10.0+cu102 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "Setup complete ✅\n" ] } ] }, { "cell_type": "markdown", "metadata": { "id": "4JnkELT0cIJg" }, "source": [ "# 1. Inference\n", "\n", "`detect.py` runs YOLOv5 inference on a variety of sources, downloading models automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases), and saving results to `runs/detect`. Example inference sources are:\n", "\n", "```shell\n", "python detect.py --source 0 # webcam\n", " img.jpg # image \n", " vid.mp4 # video\n", " path/ # directory\n", " path/*.jpg # glob\n", " 'https://youtu.be/Zgi9g1ksQHc' # YouTube\n", " 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream\n", "```" ] }, { "cell_type": "code", "metadata": { "id": "zR9ZbuQCH7FX", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "8f7e6588-215d-4ebd-93af-88b871e770a7" }, "source": [ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mdetect: \u001b[0mweights=['yolov5s.pt'], source=data/images, imgsz=[640, 640], conf_thres=0.25, iou_thres=0.45, max_det=1000, device=, view_img=False, save_txt=False, save_conf=False, save_crop=False, nosave=False, classes=None, agnostic_nms=False, augment=False, visualize=False, update=False, project=runs/detect, name=exp, exist_ok=False, line_thickness=3, hide_labels=False, hide_conf=False, half=False, dnn=False\n", "YOLOv5 🚀 v6.0-48-g84a8099 torch 1.10.0+cu102 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Fusing layers... \n", "Model Summary: 213 layers, 7225885 parameters, 0 gradients\n", "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, Done. (0.007s)\n", "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 1 tie, Done. (0.007s)\n", "Speed: 0.5ms pre-process, 6.9ms inference, 1.3ms NMS per image at shape (1, 3, 640, 640)\n", "Results saved to \u001b[1mruns/detect/exp\u001b[0m\n" ] } ] }, { "cell_type": "markdown", "metadata": { "id": "hkAzDWJ7cWTr" }, "source": [ "        \n", "" ] }, { "cell_type": "markdown", "metadata": { "id": "0eq1SMWl6Sfn" }, "source": [ "# 2. Validate\n", "Validate a model's accuracy on [COCO](https://cocodataset.org/#home) val or test-dev datasets. Models are downloaded automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases). To show results by class use the `--verbose` flag. Note that `pycocotools` metrics may be ~1% better than the equivalent repo metrics, as is visible below, due to slight differences in mAP computation." ] }, { "cell_type": "markdown", "metadata": { "id": "eyTZYGgRjnMc" }, "source": [ "## COCO val\n", "Download [COCO val 2017](https://github.com/ultralytics/yolov5/blob/74b34872fdf41941cddcf243951cdb090fbac17b/data/coco.yaml#L14) dataset (1GB - 5000 images), and test model accuracy." ] }, { "cell_type": "code", "metadata": { "id": "WQPtK1QYVaD_", "colab": { "base_uri": "https://localhost:8080/", "height": 48, "referenced_widgets": [ "eb95db7cae194218b3fcefb439b6352f", "769ecde6f2e64bacb596ce972f8d3d2d", "384a001876054c93b0af45cd1e960bfe", "dded0aeae74440f7ba2ffa0beb8dd612", "5296d28be75740b2892ae421bbec3657", "9f09facb2a6c4a7096810d327c8b551c", "25621cff5d16448cb7260e839fd0f543", "0ce7164fc0c74bb9a2b5c7037375a727", "c4c4593c10904cb5b8a5724d60c7e181", "473371611126476c88d5d42ec7031ed6", "65efdfd0d26c46e79c8c5ff3b77126cc" ] }, "outputId": "bcf9a448-1f9b-4a41-ad49-12f181faf05a" }, "source": [ "# Download COCO val\n", "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip')\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip" ], "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "eb95db7cae194218b3fcefb439b6352f", "version_minor": 0, "version_major": 2 }, "text/plain": [ " 0%| | 0.00/780M [00:00

\n", "Close the active learning loop by sampling images from your inference conditions with the `roboflow` pip package\n", "

\n", "\n", "Train a YOLOv5s model on the [COCO128](https://www.kaggle.com/ultralytics/coco128) dataset with `--data coco128.yaml`, starting from pretrained `--weights yolov5s.pt`, or from randomly initialized `--weights '' --cfg yolov5s.yaml`.\n", "\n", "- **Pretrained [Models](https://github.com/ultralytics/yolov5/tree/master/models)** are downloaded\n", "automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases)\n", "- **[Datasets](https://github.com/ultralytics/yolov5/tree/master/data)** available for autodownload include: [COCO](https://github.com/ultralytics/yolov5/blob/master/data/coco.yaml), [COCO128](https://github.com/ultralytics/yolov5/blob/master/data/coco128.yaml), [VOC](https://github.com/ultralytics/yolov5/blob/master/data/VOC.yaml), [Argoverse](https://github.com/ultralytics/yolov5/blob/master/data/Argoverse.yaml), [VisDrone](https://github.com/ultralytics/yolov5/blob/master/data/VisDrone.yaml), [GlobalWheat](https://github.com/ultralytics/yolov5/blob/master/data/GlobalWheat2020.yaml), [xView](https://github.com/ultralytics/yolov5/blob/master/data/xView.yaml), [Objects365](https://github.com/ultralytics/yolov5/blob/master/data/Objects365.yaml), [SKU-110K](https://github.com/ultralytics/yolov5/blob/master/data/SKU-110K.yaml).\n", "- **Training Results** are saved to `runs/train/` with incrementing run directories, i.e. `runs/train/exp2`, `runs/train/exp3` etc.\n", "

\n", "\n", "## Train on Custom Data with Roboflow 🌟 NEW\n", "\n", "[Roboflow](https://roboflow.com/?ref=ultralytics) enables you to easily **organize, label, and prepare** a high quality dataset with your own custom data. Roboflow also makes it easy to establish an active learning pipeline, collaborate with your team on dataset improvement, and integrate directly into your model building workflow with the `roboflow` pip package.\n", "\n", "- Custom Training Example: [https://blog.roboflow.com/how-to-train-yolov5-on-a-custom-dataset/](https://blog.roboflow.com/how-to-train-yolov5-on-a-custom-dataset/?ref=ultralytics)\n", "- Custom Training Notebook: [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/roboflow-ai/yolov5-custom-training-tutorial/blob/main/yolov5-custom-training.ipynb)\n", "
\n", "\n", "

Label images lightning fast (including with model-assisted labeling)" ] }, { "cell_type": "code", "metadata": { "id": "bOy5KI2ncnWd" }, "source": [ "# Tensorboard (optional)\n", "%load_ext tensorboard\n", "%tensorboard --logdir runs/train" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "2fLAV42oNb7M" }, "source": [ "# Weights & Biases (optional)\n", "%pip install -q wandb\n", "import wandb\n", "wandb.login()" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "1NcFxRcFdJ_O", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "8724d13d-6711-4a12-d96a-1c655e5c3549" }, "source": [ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mtrain: \u001b[0mweights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch.yaml, epochs=3, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, evolve=None, bucket=, cache=ram, image_weights=False, device=, multi_scale=False, single_cls=False, adam=False, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, linear_lr=False, label_smoothing=0.0, patience=100, freeze=0, save_period=-1, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest\n", "\u001b[34m\u001b[1mgithub: \u001b[0mup to date with https://github.com/ultralytics/yolov5 ✅\n", "YOLOv5 🚀 v6.0-48-g84a8099 torch 1.10.0+cu102 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "\u001b[34m\u001b[1mhyperparameters: \u001b[0mlr0=0.01, lrf=0.1, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0\n", "\u001b[34m\u001b[1mWeights & Biases: \u001b[0mrun 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs (RECOMMENDED)\n", "\u001b[34m\u001b[1mTensorBoard: \u001b[0mStart with 'tensorboard --logdir runs/train', view at http://localhost:6006/\n", "\n", " from n params module arguments \n", " 0 -1 1 3520 models.common.Conv [3, 32, 6, 2, 2] \n", " 1 -1 1 18560 models.common.Conv [32, 64, 3, 2] \n", " 2 -1 1 18816 models.common.C3 [64, 64, 1] \n", " 3 -1 1 73984 models.common.Conv [64, 128, 3, 2] \n", " 4 -1 2 115712 models.common.C3 [128, 128, 2] \n", " 5 -1 1 295424 models.common.Conv [128, 256, 3, 2] \n", " 6 -1 3 625152 models.common.C3 [256, 256, 3] \n", " 7 -1 1 1180672 models.common.Conv [256, 512, 3, 2] \n", " 8 -1 1 1182720 models.common.C3 [512, 512, 1] \n", " 9 -1 1 656896 models.common.SPPF [512, 512, 5] \n", " 10 -1 1 131584 models.common.Conv [512, 256, 1, 1] \n", " 11 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest'] \n", " 12 [-1, 6] 1 0 models.common.Concat [1] \n", " 13 -1 1 361984 models.common.C3 [512, 256, 1, False] \n", " 14 -1 1 33024 models.common.Conv [256, 128, 1, 1] \n", " 15 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest'] \n", " 16 [-1, 4] 1 0 models.common.Concat [1] \n", " 17 -1 1 90880 models.common.C3 [256, 128, 1, False] \n", " 18 -1 1 147712 models.common.Conv [128, 128, 3, 2] \n", " 19 [-1, 14] 1 0 models.common.Concat [1] \n", " 20 -1 1 296448 models.common.C3 [256, 256, 1, False] \n", " 21 -1 1 590336 models.common.Conv [256, 256, 3, 2] \n", " 22 [-1, 10] 1 0 models.common.Concat [1] \n", " 23 -1 1 1182720 models.common.C3 [512, 512, 1, False] \n", " 24 [17, 20, 23] 1 229245 models.yolo.Detect [80, [[10, 13, 16, 30, 33, 23], [30, 61, 62, 45, 59, 119], [116, 90, 156, 198, 373, 326]], [128, 256, 512]]\n", "Model Summary: 270 layers, 7235389 parameters, 7235389 gradients, 16.5 GFLOPs\n", "\n", "Transferred 349/349 items from yolov5s.pt\n", "Scaled weight_decay = 0.0005\n", "\u001b[34m\u001b[1moptimizer:\u001b[0m SGD with parameter groups 57 weight, 60 weight (no decay), 60 bias\n", "\u001b[34m\u001b[1malbumentations: \u001b[0mversion 1.0.3 required by YOLOv5, but version 0.1.12 is currently installed\n", "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '../datasets/coco128/labels/train2017.cache' images and labels... 128 found, 0 missing, 2 empty, 0 corrupted: 100% 128/128 [00:00\"Weights

" ] }, { "cell_type": "markdown", "metadata": { "id": "-WPvRbS5Swl6" }, "source": [ "## Local Logging\n", "\n", "All results are logged by default to `runs/train`, with a new experiment directory created for each new training as `runs/train/exp2`, `runs/train/exp3`, etc. View train and val jpgs to see mosaics, labels, predictions and augmentation effects. Note an Ultralytics **Mosaic Dataloader** is used for training (shown below), which combines 4 images into 1 mosaic during training.\n", "\n", "> \n", "`train_batch0.jpg` shows train batch 0 mosaics and labels\n", "\n", "> \n", "`test_batch0_labels.jpg` shows val batch 0 labels\n", "\n", "> \n", "`test_batch0_pred.jpg` shows val batch 0 _predictions_\n", "\n", "Training results are automatically logged to [Tensorboard](https://www.tensorflow.org/tensorboard) and [CSV](https://github.com/ultralytics/yolov5/pull/4148) as `results.csv`, which is plotted as `results.png` (below) after training completes. You can also plot any `results.csv` file manually:\n", "\n", "```python\n", "from utils.plots import plot_results \n", "plot_results('path/to/results.csv') # plot 'results.csv' as 'results.png'\n", "```\n", "\n", "\"COCO128" ] }, { "cell_type": "markdown", "metadata": { "id": "Zelyeqbyt3GD" }, "source": [ "# Environments\n", "\n", "YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including [CUDA](https://developer.nvidia.com/cuda)/[CUDNN](https://developer.nvidia.com/cudnn), [Python](https://www.python.org/) and [PyTorch](https://pytorch.org/) preinstalled):\n", "\n", "- **Google Colab and Kaggle** notebooks with free GPU: \"Open \"Open\n", "- **Google Cloud** Deep Learning VM. See [GCP Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/GCP-Quickstart)\n", "- **Amazon** Deep Learning AMI. See [AWS Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/AWS-Quickstart)\n", "- **Docker Image**. See [Docker Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/Docker-Quickstart) \"Docker\n" ] }, { "cell_type": "markdown", "metadata": { "id": "6Qu7Iesl0p54" }, "source": [ "# Status\n", "\n", "![CI CPU testing](https://github.com/ultralytics/yolov5/workflows/CI%20CPU%20testing/badge.svg)\n", "\n", "If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), testing ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on MacOS, Windows, and Ubuntu every 24 hours and on every commit.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "IEijrePND_2I" }, "source": [ "# Appendix\n", "\n", "Optional extras below. Unit tests validate repo functionality and should be run on any PRs submitted.\n" ] }, { "cell_type": "code", "metadata": { "id": "mcKoSIK2WSzj" }, "source": [ "# Reproduce\n", "for x in 'yolov5s', 'yolov5m', 'yolov5l', 'yolov5x':\n", " !python val.py --weights {x}.pt --data coco.yaml --img 640 --task speed # speed\n", " !python val.py --weights {x}.pt --data coco.yaml --img 640 --conf 0.001 --iou 0.65 # mAP" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "GMusP4OAxFu6" }, "source": [ "# PyTorch Hub\n", "import torch\n", "\n", "# Model\n", "model = torch.hub.load('ultralytics/yolov5', 'yolov5s')\n", "\n", "# Images\n", "dir = 'https://ultralytics.com/images/'\n", "imgs = [dir + f for f in ('zidane.jpg', 'bus.jpg')] # batch of images\n", "\n", "# Inference\n", "results = model(imgs)\n", "results.print() # or .show(), .save()" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "FGH0ZjkGjejy" }, "source": [ "# CI Checks\n", "%%shell\n", "export PYTHONPATH=\"$PWD\" # to run *.py. files in subdirectories\n", "rm -rf runs # remove runs/\n", "for m in yolov5n; do # models\n", " python train.py --img 64 --batch 32 --weights $m.pt --epochs 1 --device 0 # train pretrained\n", " python train.py --img 64 --batch 32 --weights '' --cfg $m.yaml --epochs 1 --device 0 # train scratch\n", " for d in 0 cpu; do # devices\n", " python val.py --weights $m.pt --device $d # val official\n", " python val.py --weights runs/train/exp/weights/best.pt --device $d # val custom\n", " python detect.py --weights $m.pt --device $d # detect official\n", " python detect.py --weights runs/train/exp/weights/best.pt --device $d # detect custom\n", " done\n", " python hubconf.py # hub\n", " python models/yolo.py --cfg $m.yaml # build PyTorch model\n", " python models/tf.py --weights $m.pt # build TensorFlow model\n", " python export.py --img 64 --batch 1 --weights $m.pt --include torchscript onnx # export\n", "done" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "gogI-kwi3Tye" }, "source": [ "# Profile\n", "from utils.torch_utils import profile\n", "\n", "m1 = lambda x: x * torch.sigmoid(x)\n", "m2 = torch.nn.SiLU()\n", "results = profile(input=torch.randn(16, 3, 640, 640), ops=[m1, m2], n=100)" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "RVRSOhEvUdb5" }, "source": [ "# Evolve\n", "!python train.py --img 640 --batch 64 --epochs 100 --data coco128.yaml --weights yolov5s.pt --cache --noautoanchor --evolve\n", "!d=runs/train/evolve && cp evolve.* $d && zip -r evolve.zip $d && gsutil mv evolve.zip gs://bucket # upload results (optional)" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "BSgFCAcMbk1R" }, "source": [ "# VOC\n", "for b, m in zip([64, 48, 32, 16], ['yolov5s', 'yolov5m', 'yolov5l', 'yolov5x']): # zip(batch_size, model)\n", " !python train.py --batch {b} --weights {m}.pt --data VOC.yaml --epochs 50 --cache --img 512 --nosave --hyp hyp.finetune.yaml --project VOC --name {m}" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "VTRwsvA9u7ln" }, "source": [ "# TensorRT \n", "# https://developer.nvidia.com/nvidia-tensorrt-download\n", "!lsb_release -a # check system\n", "%ls /usr/local | grep cuda # check CUDA\n", "!wget https://ultralytics.com/assets/TensorRT-8.2.0.6.Linux.x86_64-gnu.cuda-11.4.cudnn8.2.tar.gz # download\n", "![ -d /content/TensorRT-8.2.0.6/ ] || tar -C /content/ -zxf ./TensorRT-8.2.0.6.Linux.x86_64-gnu.cuda-11.4.cudnn8.2.tar.gz # unzip\n", "%pip list | grep tensorrt || pip install /content/TensorRT-8.2.0.6/python/tensorrt-8.2.0.6-cp37-none-linux_x86_64.whl # install\n", "%env LD_LIBRARY_PATH=/usr/local/cuda-11.1/lib64:/content/cuda-11.1/lib64:/content/TensorRT-8.2.0.6/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64 # add to path\n", "\n", "!python export.py --weights yolov5s.pt --include engine --imgsz 640 640 --device 0\n", "!python detect.py --weights yolov5s.engine --imgsz 640 640 --device 0" ], "execution_count": null, "outputs": [] } ] } ================================================ FILE: utils/__init__.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ utils/initialization """ def notebook_init(verbose=True): # Check system software and hardware print('Checking setup...') import os import shutil from utils.general import check_requirements, emojis, is_colab from utils.torch_utils import select_device # imports check_requirements(('psutil', 'IPython')) import psutil from IPython import display # to display images and clear console output if is_colab(): shutil.rmtree('/content/sample_data', ignore_errors=True) # remove colab /sample_data directory if verbose: # System info # gb = 1 / 1000 ** 3 # bytes to GB gib = 1 / 1024 ** 3 # bytes to GiB ram = psutil.virtual_memory().total total, used, free = shutil.disk_usage("/") display.clear_output() s = f'({os.cpu_count()} CPUs, {ram * gib:.1f} GB RAM, {(total - free) * gib:.1f}/{total * gib:.1f} GB disk)' else: s = '' select_device(newline=False) print(emojis(f'Setup complete ✅ {s}')) return display ================================================ FILE: utils/activations.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Activation functions """ import torch import torch.nn as nn import torch.nn.functional as F # SiLU https://arxiv.org/pdf/1606.08415.pdf ---------------------------------------------------------------------------- class SiLU(nn.Module): # export-friendly version of nn.SiLU() @staticmethod def forward(x): return x * torch.sigmoid(x) class Hardswish(nn.Module): # export-friendly version of nn.Hardswish() @staticmethod def forward(x): # return x * F.hardsigmoid(x) # for TorchScript and CoreML return x * F.hardtanh(x + 3, 0.0, 6.0) / 6.0 # for TorchScript, CoreML and ONNX # Mish https://github.com/digantamisra98/Mish -------------------------------------------------------------------------- class Mish(nn.Module): @staticmethod def forward(x): return x * F.softplus(x).tanh() class MemoryEfficientMish(nn.Module): class F(torch.autograd.Function): @staticmethod def forward(ctx, x): ctx.save_for_backward(x) return x.mul(torch.tanh(F.softplus(x))) # x * tanh(ln(1 + exp(x))) @staticmethod def backward(ctx, grad_output): x = ctx.saved_tensors[0] sx = torch.sigmoid(x) fx = F.softplus(x).tanh() return grad_output * (fx + x * sx * (1 - fx * fx)) def forward(self, x): return self.F.apply(x) # FReLU https://arxiv.org/abs/2007.11824 ------------------------------------------------------------------------------- class FReLU(nn.Module): def __init__(self, c1, k=3): # ch_in, kernel super().__init__() self.conv = nn.Conv2d(c1, c1, k, 1, 1, groups=c1, bias=False) self.bn = nn.BatchNorm2d(c1) def forward(self, x): return torch.max(x, self.bn(self.conv(x))) # ACON https://arxiv.org/pdf/2009.04759.pdf ---------------------------------------------------------------------------- class AconC(nn.Module): r""" ACON activation (activate or not). AconC: (p1*x-p2*x) * sigmoid(beta*(p1*x-p2*x)) + p2*x, beta is a learnable parameter according to "Activate or Not: Learning Customized Activation" . """ def __init__(self, c1): super().__init__() self.p1 = nn.Parameter(torch.randn(1, c1, 1, 1)) self.p2 = nn.Parameter(torch.randn(1, c1, 1, 1)) self.beta = nn.Parameter(torch.ones(1, c1, 1, 1)) def forward(self, x): dpx = (self.p1 - self.p2) * x return dpx * torch.sigmoid(self.beta * dpx) + self.p2 * x class MetaAconC(nn.Module): r""" ACON activation (activate or not). MetaAconC: (p1*x-p2*x) * sigmoid(beta*(p1*x-p2*x)) + p2*x, beta is generated by a small network according to "Activate or Not: Learning Customized Activation" . """ def __init__(self, c1, k=1, s=1, r=16): # ch_in, kernel, stride, r super().__init__() c2 = max(r, c1 // r) self.p1 = nn.Parameter(torch.randn(1, c1, 1, 1)) self.p2 = nn.Parameter(torch.randn(1, c1, 1, 1)) self.fc1 = nn.Conv2d(c1, c2, k, s, bias=True) self.fc2 = nn.Conv2d(c2, c1, k, s, bias=True) # self.bn1 = nn.BatchNorm2d(c2) # self.bn2 = nn.BatchNorm2d(c1) def forward(self, x): y = x.mean(dim=2, keepdims=True).mean(dim=3, keepdims=True) # batch-size 1 bug/instabilities https://github.com/ultralytics/yolov5/issues/2891 # beta = torch.sigmoid(self.bn2(self.fc2(self.bn1(self.fc1(y))))) # bug/unstable beta = torch.sigmoid(self.fc2(self.fc1(y))) # bug patch BN layers removed dpx = (self.p1 - self.p2) * x return dpx * torch.sigmoid(beta * dpx) + self.p2 * x ================================================ FILE: utils/augmentations.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Image augmentation functions """ import math import random import cv2 import numpy as np from utils.general import LOGGER, check_version, colorstr, resample_segments, segment2box from utils.metrics import bbox_ioa class Albumentations: # YOLOv5 Albumentations class (optional, only used if package is installed) def __init__(self): self.transform = None try: import albumentations as A check_version(A.__version__, '1.0.3', hard=True) # version requirement self.transform = A.Compose([ A.Blur(p=0.01), A.MedianBlur(p=0.01), A.ToGray(p=0.01), A.CLAHE(p=0.01), A.RandomBrightnessContrast(p=0.0), A.RandomGamma(p=0.0), A.ImageCompression(quality_lower=75, p=0.0)], bbox_params=A.BboxParams(format='yolo', label_fields=['class_labels'])) LOGGER.info(colorstr('albumentations: ') + ', '.join(f'{x}' for x in self.transform.transforms if x.p)) except ImportError: # package not installed, skip pass except Exception as e: LOGGER.info(colorstr('albumentations: ') + f'{e}') def __call__(self, im, labels, p=1.0): if self.transform and random.random() < p: new = self.transform(image=im, bboxes=labels[:, 1:], class_labels=labels[:, 0]) # transformed im, labels = new['image'], np.array([[c, *b] for c, b in zip(new['class_labels'], new['bboxes'])]) return im, labels def augment_hsv(im, hgain=0.5, sgain=0.5, vgain=0.5): # HSV color-space augmentation if hgain or sgain or vgain: r = np.random.uniform(-1, 1, 3) * [hgain, sgain, vgain] + 1 # random gains hue, sat, val = cv2.split(cv2.cvtColor(im, cv2.COLOR_BGR2HSV)) dtype = im.dtype # uint8 x = np.arange(0, 256, dtype=r.dtype) lut_hue = ((x * r[0]) % 180).astype(dtype) lut_sat = np.clip(x * r[1], 0, 255).astype(dtype) lut_val = np.clip(x * r[2], 0, 255).astype(dtype) im_hsv = cv2.merge((cv2.LUT(hue, lut_hue), cv2.LUT(sat, lut_sat), cv2.LUT(val, lut_val))) cv2.cvtColor(im_hsv, cv2.COLOR_HSV2BGR, dst=im) # no return needed def hist_equalize(im, clahe=True, bgr=False): # Equalize histogram on BGR image 'im' with im.shape(n,m,3) and range 0-255 yuv = cv2.cvtColor(im, cv2.COLOR_BGR2YUV if bgr else cv2.COLOR_RGB2YUV) if clahe: c = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8)) yuv[:, :, 0] = c.apply(yuv[:, :, 0]) else: yuv[:, :, 0] = cv2.equalizeHist(yuv[:, :, 0]) # equalize Y channel histogram return cv2.cvtColor(yuv, cv2.COLOR_YUV2BGR if bgr else cv2.COLOR_YUV2RGB) # convert YUV image to RGB def replicate(im, labels): # Replicate labels h, w = im.shape[:2] boxes = labels[:, 1:].astype(int) x1, y1, x2, y2 = boxes.T s = ((x2 - x1) + (y2 - y1)) / 2 # side length (pixels) for i in s.argsort()[:round(s.size * 0.5)]: # smallest indices x1b, y1b, x2b, y2b = boxes[i] bh, bw = y2b - y1b, x2b - x1b yc, xc = int(random.uniform(0, h - bh)), int(random.uniform(0, w - bw)) # offset x, y x1a, y1a, x2a, y2a = [xc, yc, xc + bw, yc + bh] im[y1a:y2a, x1a:x2a] = im[y1b:y2b, x1b:x2b] # im4[ymin:ymax, xmin:xmax] labels = np.append(labels, [[labels[i, 0], x1a, y1a, x2a, y2a]], axis=0) return im, labels def letterbox(im, new_shape=(640, 640), color=(114, 114, 114), auto=True, scaleFill=False, scaleup=True, stride=32): # Resize and pad image while meeting stride-multiple constraints shape = im.shape[:2] # current shape [height, width] if isinstance(new_shape, int): new_shape = (new_shape, new_shape) # Scale ratio (new / old) r = min(new_shape[0] / shape[0], new_shape[1] / shape[1]) if not scaleup: # only scale down, do not scale up (for better val mAP) r = min(r, 1.0) # Compute padding ratio = r, r # width, height ratios new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r)) dw, dh = new_shape[1] - new_unpad[0], new_shape[0] - new_unpad[1] # wh padding if auto: # minimum rectangle dw, dh = np.mod(dw, stride), np.mod(dh, stride) # wh padding elif scaleFill: # stretch dw, dh = 0.0, 0.0 new_unpad = (new_shape[1], new_shape[0]) ratio = new_shape[1] / shape[1], new_shape[0] / shape[0] # width, height ratios dw /= 2 # divide padding into 2 sides dh /= 2 if shape[::-1] != new_unpad: # resize im = cv2.resize(im, new_unpad, interpolation=cv2.INTER_LINEAR) top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1)) left, right = int(round(dw - 0.1)), int(round(dw + 0.1)) im = cv2.copyMakeBorder(im, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color) # add border return im, ratio, (dw, dh) def random_perspective(im, targets=(), segments=(), degrees=10, translate=.1, scale=.1, shear=10, perspective=0.0, border=(0, 0)): # torchvision.transforms.RandomAffine(degrees=(-10, 10), translate=(0.1, 0.1), scale=(0.9, 1.1), shear=(-10, 10)) # targets = [cls, xyxy] height = im.shape[0] + border[0] * 2 # shape(h,w,c) width = im.shape[1] + border[1] * 2 # Center C = np.eye(3) C[0, 2] = -im.shape[1] / 2 # x translation (pixels) C[1, 2] = -im.shape[0] / 2 # y translation (pixels) # Perspective P = np.eye(3) P[2, 0] = random.uniform(-perspective, perspective) # x perspective (about y) P[2, 1] = random.uniform(-perspective, perspective) # y perspective (about x) # Rotation and Scale R = np.eye(3) a = random.uniform(-degrees, degrees) # a += random.choice([-180, -90, 0, 90]) # add 90deg rotations to small rotations s = random.uniform(1 - scale, 1 + scale) # s = 2 ** random.uniform(-scale, scale) R[:2] = cv2.getRotationMatrix2D(angle=a, center=(0, 0), scale=s) # Shear S = np.eye(3) S[0, 1] = math.tan(random.uniform(-shear, shear) * math.pi / 180) # x shear (deg) S[1, 0] = math.tan(random.uniform(-shear, shear) * math.pi / 180) # y shear (deg) # Translation T = np.eye(3) T[0, 2] = random.uniform(0.5 - translate, 0.5 + translate) * width # x translation (pixels) T[1, 2] = random.uniform(0.5 - translate, 0.5 + translate) * height # y translation (pixels) # Combined rotation matrix M = T @ S @ R @ P @ C # order of operations (right to left) is IMPORTANT if (border[0] != 0) or (border[1] != 0) or (M != np.eye(3)).any(): # image changed if perspective: im = cv2.warpPerspective(im, M, dsize=(width, height), borderValue=(114, 114, 114)) else: # affine im = cv2.warpAffine(im, M[:2], dsize=(width, height), borderValue=(114, 114, 114)) # Visualize # import matplotlib.pyplot as plt # ax = plt.subplots(1, 2, figsize=(12, 6))[1].ravel() # ax[0].imshow(im[:, :, ::-1]) # base # ax[1].imshow(im2[:, :, ::-1]) # warped # Transform label coordinates n = len(targets) if n: use_segments = any(x.any() for x in segments) new = np.zeros((n, 4)) if use_segments: # warp segments segments = resample_segments(segments) # upsample for i, segment in enumerate(segments): xy = np.ones((len(segment), 3)) xy[:, :2] = segment xy = xy @ M.T # transform xy = xy[:, :2] / xy[:, 2:3] if perspective else xy[:, :2] # perspective rescale or affine # clip new[i] = segment2box(xy, width, height) else: # warp boxes xy = np.ones((n * 4, 3)) xy[:, :2] = targets[:, [1, 2, 3, 4, 1, 4, 3, 2]].reshape(n * 4, 2) # x1y1, x2y2, x1y2, x2y1 xy = xy @ M.T # transform xy = (xy[:, :2] / xy[:, 2:3] if perspective else xy[:, :2]).reshape(n, 8) # perspective rescale or affine # create new boxes x = xy[:, [0, 2, 4, 6]] y = xy[:, [1, 3, 5, 7]] new = np.concatenate((x.min(1), y.min(1), x.max(1), y.max(1))).reshape(4, n).T # clip new[:, [0, 2]] = new[:, [0, 2]].clip(0, width) new[:, [1, 3]] = new[:, [1, 3]].clip(0, height) # filter candidates i = box_candidates(box1=targets[:, 1:5].T * s, box2=new.T, area_thr=0.01 if use_segments else 0.10) targets = targets[i] targets[:, 1:5] = new[i] return im, targets def copy_paste(im, labels, segments, p=0.5): # Implement Copy-Paste augmentation https://arxiv.org/abs/2012.07177, labels as nx5 np.array(cls, xyxy) n = len(segments) if p and n: h, w, c = im.shape # height, width, channels im_new = np.zeros(im.shape, np.uint8) for j in random.sample(range(n), k=round(p * n)): l, s = labels[j], segments[j] box = w - l[3], l[2], w - l[1], l[4] ioa = bbox_ioa(box, labels[:, 1:5]) # intersection over area if (ioa < 0.30).all(): # allow 30% obscuration of existing labels labels = np.concatenate((labels, [[l[0], *box]]), 0) segments.append(np.concatenate((w - s[:, 0:1], s[:, 1:2]), 1)) cv2.drawContours(im_new, [segments[j].astype(np.int32)], -1, (255, 255, 255), cv2.FILLED) result = cv2.bitwise_and(src1=im, src2=im_new) result = cv2.flip(result, 1) # augment segments (flip left-right) i = result > 0 # pixels to replace # i[:, :] = result.max(2).reshape(h, w, 1) # act over ch im[i] = result[i] # cv2.imwrite('debug.jpg', im) # debug return im, labels, segments def cutout(im, labels, p=0.5): # Applies image cutout augmentation https://arxiv.org/abs/1708.04552 if random.random() < p: h, w = im.shape[:2] scales = [0.5] * 1 + [0.25] * 2 + [0.125] * 4 + [0.0625] * 8 + [0.03125] * 16 # image size fraction for s in scales: mask_h = random.randint(1, int(h * s)) # create random masks mask_w = random.randint(1, int(w * s)) # box xmin = max(0, random.randint(0, w) - mask_w // 2) ymin = max(0, random.randint(0, h) - mask_h // 2) xmax = min(w, xmin + mask_w) ymax = min(h, ymin + mask_h) # apply random color mask im[ymin:ymax, xmin:xmax] = [random.randint(64, 191) for _ in range(3)] # return unobscured labels if len(labels) and s > 0.03: box = np.array([xmin, ymin, xmax, ymax], dtype=np.float32) ioa = bbox_ioa(box, labels[:, 1:5]) # intersection over area labels = labels[ioa < 0.60] # remove >60% obscured labels return labels def mixup(im, labels, im2, labels2): # Applies MixUp augmentation https://arxiv.org/pdf/1710.09412.pdf r = np.random.beta(32.0, 32.0) # mixup ratio, alpha=beta=32.0 im = (im * r + im2 * (1 - r)).astype(np.uint8) labels = np.concatenate((labels, labels2), 0) return im, labels def box_candidates(box1, box2, wh_thr=2, ar_thr=20, area_thr=0.1, eps=1e-16): # box1(4,n), box2(4,n) # Compute candidate boxes: box1 before augment, box2 after augment, wh_thr (pixels), aspect_ratio_thr, area_ratio w1, h1 = box1[2] - box1[0], box1[3] - box1[1] w2, h2 = box2[2] - box2[0], box2[3] - box2[1] ar = np.maximum(w2 / (h2 + eps), h2 / (w2 + eps)) # aspect ratio return (w2 > wh_thr) & (h2 > wh_thr) & (w2 * h2 / (w1 * h1 + eps) > area_thr) & (ar < ar_thr) # candidates ================================================ FILE: utils/autoanchor.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Auto-anchor utils """ import random import numpy as np import torch import yaml from tqdm import tqdm from utils.general import LOGGER, colorstr, emojis PREFIX = colorstr('AutoAnchor: ') def check_anchor_order(m): # Check anchor order against stride order for YOLOv5 Detect() module m, and correct if necessary a = m.anchors.prod(-1).view(-1) # anchor area da = a[-1] - a[0] # delta a ds = m.stride[-1] - m.stride[0] # delta s if da.sign() != ds.sign(): # same order LOGGER.info(f'{PREFIX}Reversing anchor order') m.anchors[:] = m.anchors.flip(0) def check_anchors(dataset, model, thr=4.0, imgsz=640): # Check anchor fit to data, recompute if necessary m = model.module.model[-1] if hasattr(model, 'module') else model.model[-1] # Detect() shapes = imgsz * dataset.shapes / dataset.shapes.max(1, keepdims=True) scale = np.random.uniform(0.9, 1.1, size=(shapes.shape[0], 1)) # augment scale wh = torch.tensor(np.concatenate([l[:, 3:5] * s for s, l in zip(shapes * scale, dataset.labels)])).float() # wh def metric(k): # compute metric r = wh[:, None] / k[None] x = torch.min(r, 1 / r).min(2)[0] # ratio metric best = x.max(1)[0] # best_x aat = (x > 1 / thr).float().sum(1).mean() # anchors above threshold bpr = (best > 1 / thr).float().mean() # best possible recall return bpr, aat anchors = m.anchors.clone() * m.stride.to(m.anchors.device).view(-1, 1, 1) # current anchors bpr, aat = metric(anchors.cpu().view(-1, 2)) s = f'\n{PREFIX}{aat:.2f} anchors/target, {bpr:.3f} Best Possible Recall (BPR). ' if bpr > 0.98: # threshold to recompute LOGGER.info(emojis(f'{s}Current anchors are a good fit to dataset ✅')) else: LOGGER.info(emojis(f'{s}Anchors are a poor fit to dataset ⚠️, attempting to improve...')) na = m.anchors.numel() // 2 # number of anchors try: anchors = kmean_anchors(dataset, n=na, img_size=imgsz, thr=thr, gen=1000, verbose=False) except Exception as e: LOGGER.info(f'{PREFIX}ERROR: {e}') new_bpr = metric(anchors)[0] if new_bpr > bpr: # replace anchors anchors = torch.tensor(anchors, device=m.anchors.device).type_as(m.anchors) m.anchors[:] = anchors.clone().view_as(m.anchors) / m.stride.to(m.anchors.device).view(-1, 1, 1) # loss check_anchor_order(m) LOGGER.info(f'{PREFIX}New anchors saved to model. Update model *.yaml to use these anchors in the future.') else: LOGGER.info(f'{PREFIX}Original anchors better than new anchors. Proceeding with original anchors.') def kmean_anchors(dataset='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=1000, verbose=True): """ Creates kmeans-evolved anchors from training dataset Arguments: dataset: path to data.yaml, or a loaded dataset n: number of anchors img_size: image size used for training thr: anchor-label wh ratio threshold hyperparameter hyp['anchor_t'] used for training, default=4.0 gen: generations to evolve anchors using genetic algorithm verbose: print all results Return: k: kmeans evolved anchors Usage: from utils.autoanchor import *; _ = kmean_anchors() """ from scipy.cluster.vq import kmeans thr = 1 / thr def metric(k, wh): # compute metrics r = wh[:, None] / k[None] x = torch.min(r, 1 / r).min(2)[0] # ratio metric # x = wh_iou(wh, torch.tensor(k)) # iou metric return x, x.max(1)[0] # x, best_x def anchor_fitness(k): # mutation fitness _, best = metric(torch.tensor(k, dtype=torch.float32), wh) return (best * (best > thr).float()).mean() # fitness def print_results(k, verbose=True): k = k[np.argsort(k.prod(1))] # sort small to large x, best = metric(k, wh0) bpr, aat = (best > thr).float().mean(), (x > thr).float().mean() * n # best possible recall, anch > thr s = f'{PREFIX}thr={thr:.2f}: {bpr:.4f} best possible recall, {aat:.2f} anchors past thr\n' \ f'{PREFIX}n={n}, img_size={img_size}, metric_all={x.mean():.3f}/{best.mean():.3f}-mean/best, ' \ f'past_thr={x[x > thr].mean():.3f}-mean: ' for i, x in enumerate(k): s += '%i,%i, ' % (round(x[0]), round(x[1])) if verbose: LOGGER.info(s[:-2]) return k if isinstance(dataset, str): # *.yaml file with open(dataset, errors='ignore') as f: data_dict = yaml.safe_load(f) # model dict from utils.datasets import LoadImagesAndLabels dataset = LoadImagesAndLabels(data_dict['train'], augment=True, rect=True) # Get label wh shapes = img_size * dataset.shapes / dataset.shapes.max(1, keepdims=True) wh0 = np.concatenate([l[:, 3:5] * s for s, l in zip(shapes, dataset.labels)]) # wh # Filter i = (wh0 < 3.0).any(1).sum() if i: LOGGER.info(f'{PREFIX}WARNING: Extremely small objects found. {i} of {len(wh0)} labels are < 3 pixels in size.') wh = wh0[(wh0 >= 2.0).any(1)] # filter > 2 pixels # wh = wh * (np.random.rand(wh.shape[0], 1) * 0.9 + 0.1) # multiply by random scale 0-1 # Kmeans calculation LOGGER.info(f'{PREFIX}Running kmeans for {n} anchors on {len(wh)} points...') s = wh.std(0) # sigmas for whitening k, dist = kmeans(wh / s, n, iter=30) # points, mean distance assert len(k) == n, f'{PREFIX}ERROR: scipy.cluster.vq.kmeans requested {n} points but returned only {len(k)}' k *= s wh = torch.tensor(wh, dtype=torch.float32) # filtered wh0 = torch.tensor(wh0, dtype=torch.float32) # unfiltered k = print_results(k, verbose=False) # Plot # k, d = [None] * 20, [None] * 20 # for i in tqdm(range(1, 21)): # k[i-1], d[i-1] = kmeans(wh / s, i) # points, mean distance # fig, ax = plt.subplots(1, 2, figsize=(14, 7), tight_layout=True) # ax = ax.ravel() # ax[0].plot(np.arange(1, 21), np.array(d) ** 2, marker='.') # fig, ax = plt.subplots(1, 2, figsize=(14, 7)) # plot wh # ax[0].hist(wh[wh[:, 0]<100, 0],400) # ax[1].hist(wh[wh[:, 1]<100, 1],400) # fig.savefig('wh.png', dpi=200) # Evolve npr = np.random f, sh, mp, s = anchor_fitness(k), k.shape, 0.9, 0.1 # fitness, generations, mutation prob, sigma pbar = tqdm(range(gen), desc=f'{PREFIX}Evolving anchors with Genetic Algorithm:') # progress bar for _ in pbar: v = np.ones(sh) while (v == 1).all(): # mutate until a change occurs (prevent duplicates) v = ((npr.random(sh) < mp) * random.random() * npr.randn(*sh) * s + 1).clip(0.3, 3.0) kg = (k.copy() * v).clip(min=2.0) fg = anchor_fitness(kg) if fg > f: f, k = fg, kg.copy() pbar.desc = f'{PREFIX}Evolving anchors with Genetic Algorithm: fitness = {f:.4f}' if verbose: print_results(k, verbose) return print_results(k) ================================================ FILE: utils/autobatch.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Auto-batch utils """ from copy import deepcopy import numpy as np import torch from torch.cuda import amp from utils.general import LOGGER, colorstr from utils.torch_utils import profile def check_train_batch_size(model, imgsz=640): # Check YOLOv5 training batch size with amp.autocast(): return autobatch(deepcopy(model).train(), imgsz) # compute optimal batch size def autobatch(model, imgsz=640, fraction=0.9, batch_size=16): # Automatically estimate best batch size to use `fraction` of available CUDA memory # Usage: # import torch # from utils.autobatch import autobatch # model = torch.hub.load('ultralytics/yolov5', 'yolov5s', autoshape=False) # print(autobatch(model)) prefix = colorstr('AutoBatch: ') LOGGER.info(f'{prefix}Computing optimal batch size for --imgsz {imgsz}') device = next(model.parameters()).device # get model device if device.type == 'cpu': LOGGER.info(f'{prefix}CUDA not detected, using default CPU batch-size {batch_size}') return batch_size d = str(device).upper() # 'CUDA:0' properties = torch.cuda.get_device_properties(device) # device properties t = properties.total_memory / 1024 ** 3 # (GiB) r = torch.cuda.memory_reserved(device) / 1024 ** 3 # (GiB) a = torch.cuda.memory_allocated(device) / 1024 ** 3 # (GiB) f = t - (r + a) # free inside reserved LOGGER.info(f'{prefix}{d} ({properties.name}) {t:.2f}G total, {r:.2f}G reserved, {a:.2f}G allocated, {f:.2f}G free') batch_sizes = [1, 2, 4, 8, 16] try: img = [torch.zeros(b, 3, imgsz, imgsz) for b in batch_sizes] y = profile(img, model, n=3, device=device) except Exception as e: LOGGER.warning(f'{prefix}{e}') y = [x[2] for x in y if x] # memory [2] batch_sizes = batch_sizes[:len(y)] p = np.polyfit(batch_sizes, y, deg=1) # first degree polynomial fit b = int((f * fraction - p[1]) / p[0]) # y intercept (optimal batch size) LOGGER.info(f'{prefix}Using batch-size {b} for {d} {t * fraction:.2f}G/{t:.2f}G ({fraction * 100:.0f}%)') return b ================================================ FILE: utils/aws/__init__.py ================================================ ================================================ FILE: utils/aws/mime.sh ================================================ # AWS EC2 instance startup 'MIME' script https://aws.amazon.com/premiumsupport/knowledge-center/execute-user-data-ec2/ # This script will run on every instance restart, not only on first start # --- DO NOT COPY ABOVE COMMENTS WHEN PASTING INTO USERDATA --- Content-Type: multipart/mixed; boundary="//" MIME-Version: 1.0 --// Content-Type: text/cloud-config; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="cloud-config.txt" #cloud-config cloud_final_modules: - [scripts-user, always] --// Content-Type: text/x-shellscript; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="userdata.txt" #!/bin/bash # --- paste contents of userdata.sh here --- --// ================================================ FILE: utils/aws/resume.py ================================================ # Resume all interrupted trainings in yolov5/ dir including DDP trainings # Usage: $ python utils/aws/resume.py import os import sys from pathlib import Path import torch import yaml FILE = Path(__file__).resolve() ROOT = FILE.parents[2] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH port = 0 # --master_port path = Path('').resolve() for last in path.rglob('*/**/last.pt'): ckpt = torch.load(last) if ckpt['optimizer'] is None: continue # Load opt.yaml with open(last.parent.parent / 'opt.yaml', errors='ignore') as f: opt = yaml.safe_load(f) # Get device count d = opt['device'].split(',') # devices nd = len(d) # number of devices ddp = nd > 1 or (nd == 0 and torch.cuda.device_count() > 1) # distributed data parallel if ddp: # multi-GPU port += 1 cmd = f'python -m torch.distributed.run --nproc_per_node {nd} --master_port {port} train.py --resume {last}' else: # single-GPU cmd = f'python train.py --resume {last}' cmd += ' > /dev/null 2>&1 &' # redirect output to dev/null and run in daemon thread print(cmd) os.system(cmd) ================================================ FILE: utils/aws/userdata.sh ================================================ #!/bin/bash # AWS EC2 instance startup script https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/user-data.html # This script will run only once on first instance start (for a re-start script see mime.sh) # /home/ubuntu (ubuntu) or /home/ec2-user (amazon-linux) is working dir # Use >300 GB SSD cd home/ubuntu if [ ! -d yolov5 ]; then echo "Running first-time script." # install dependencies, download COCO, pull Docker git clone https://github.com/ultralytics/yolov5 -b master && sudo chmod -R 777 yolov5 cd yolov5 bash data/scripts/get_coco.sh && echo "COCO done." & sudo docker pull ultralytics/yolov5:latest && echo "Docker done." & python -m pip install --upgrade pip && pip install -r requirements.txt && python detect.py && echo "Requirements done." & wait && echo "All tasks done." # finish background tasks else echo "Running re-start script." # resume interrupted runs i=0 list=$(sudo docker ps -qa) # container list i.e. $'one\ntwo\nthree\nfour' while IFS= read -r id; do ((i++)) echo "restarting container $i: $id" sudo docker start $id # sudo docker exec -it $id python train.py --resume # single-GPU sudo docker exec -d $id python utils/aws/resume.py # multi-scenario done <<<"$list" fi ================================================ FILE: utils/callbacks.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Callback utils """ class Callbacks: """" Handles all registered callbacks for YOLOv5 Hooks """ def __init__(self): # Define the available callbacks self._callbacks = { 'on_pretrain_routine_start': [], 'on_pretrain_routine_end': [], 'on_train_start': [], 'on_train_epoch_start': [], 'on_train_batch_start': [], 'optimizer_step': [], 'on_before_zero_grad': [], 'on_train_batch_end': [], 'on_train_epoch_end': [], 'on_val_start': [], 'on_val_batch_start': [], 'on_val_image_end': [], 'on_val_batch_end': [], 'on_val_end': [], 'on_fit_epoch_end': [], # fit = train + val 'on_model_save': [], 'on_train_end': [], 'teardown': [], } def register_action(self, hook, name='', callback=None): """ Register a new action to a callback hook Args: hook The callback hook name to register the action to name The name of the action for later reference callback The callback to fire """ assert hook in self._callbacks, f"hook '{hook}' not found in callbacks {self._callbacks}" assert callable(callback), f"callback '{callback}' is not callable" self._callbacks[hook].append({'name': name, 'callback': callback}) def get_registered_actions(self, hook=None): """" Returns all the registered actions by callback hook Args: hook The name of the hook to check, defaults to all """ if hook: return self._callbacks[hook] else: return self._callbacks def run(self, hook, *args, **kwargs): """ Loop through the registered actions and fire all callbacks Args: hook The name of the hook to check, defaults to all args Arguments to receive from YOLOv5 kwargs Keyword Arguments to receive from YOLOv5 """ assert hook in self._callbacks, f"hook '{hook}' not found in callbacks {self._callbacks}" for logger in self._callbacks[hook]: logger['callback'](*args, **kwargs) ================================================ FILE: utils/datasets.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Dataloaders and dataset utils """ import glob import hashlib import json import os import random import shutil import time from itertools import repeat from multiprocessing.pool import Pool, ThreadPool from pathlib import Path from threading import Thread from zipfile import ZipFile import cv2 import numpy as np import torch import torch.nn.functional as F import yaml from PIL import ExifTags, Image, ImageOps from torch.utils.data import DataLoader, Dataset, dataloader, distributed from tqdm import tqdm from utils.augmentations import Albumentations, augment_hsv, copy_paste, letterbox, mixup, random_perspective from utils.general import (LOGGER, NUM_THREADS, check_dataset, check_requirements, check_yaml, clean_str, segments2boxes, xyn2xy, xywh2xyxy, xywhn2xyxy, xyxy2xywhn) from utils.torch_utils import torch_distributed_zero_first # Parameters HELP_URL = 'https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' IMG_FORMATS = ['bmp', 'jpg', 'jpeg', 'png', 'tif', 'tiff', 'dng', 'webp', 'mpo'] # acceptable image suffixes VID_FORMATS = ['mov', 'avi', 'mp4', 'mpg', 'mpeg', 'm4v', 'wmv', 'mkv'] # acceptable video suffixes WORLD_SIZE = int(os.getenv('WORLD_SIZE', 1)) # DPP # Get orientation exif tag for orientation in ExifTags.TAGS.keys(): if ExifTags.TAGS[orientation] == 'Orientation': break def get_hash(paths): # Returns a single hash value of a list of paths (files or dirs) size = sum(os.path.getsize(p) for p in paths if os.path.exists(p)) # sizes h = hashlib.md5(str(size).encode()) # hash sizes h.update(''.join(paths).encode()) # hash paths return h.hexdigest() # return hash def exif_size(img): # Returns exif-corrected PIL size s = img.size # (width, height) try: rotation = dict(img._getexif().items())[orientation] if rotation == 6: # rotation 270 s = (s[1], s[0]) elif rotation == 8: # rotation 90 s = (s[1], s[0]) except: pass return s def exif_transpose(image): """ Transpose a PIL image accordingly if it has an EXIF Orientation tag. Inplace version of https://github.com/python-pillow/Pillow/blob/master/src/PIL/ImageOps.py exif_transpose() :param image: The image to transpose. :return: An image. """ exif = image.getexif() orientation = exif.get(0x0112, 1) # default 1 if orientation > 1: method = {2: Image.FLIP_LEFT_RIGHT, 3: Image.ROTATE_180, 4: Image.FLIP_TOP_BOTTOM, 5: Image.TRANSPOSE, 6: Image.ROTATE_270, 7: Image.TRANSVERSE, 8: Image.ROTATE_90, }.get(orientation) if method is not None: image = image.transpose(method) del exif[0x0112] image.info["exif"] = exif.tobytes() return image def create_dataloader(path, imgsz, batch_size, stride, single_cls=False, hyp=None, augment=False, cache=False, pad=0.0, rect=False, rank=-1, workers=8, image_weights=False, quad=False, prefix='', shuffle=False): if rect and shuffle: LOGGER.warning('WARNING: --rect is incompatible with DataLoader shuffle, setting shuffle=False') shuffle = False with torch_distributed_zero_first(rank): # init dataset *.cache only once if DDP dataset = LoadImagesAndLabels(path, imgsz, batch_size, augment=augment, # augmentation hyp=hyp, # hyperparameters rect=rect, # rectangular batches cache_images=cache, single_cls=single_cls, stride=int(stride), pad=pad, image_weights=image_weights, prefix=prefix) batch_size = min(batch_size, len(dataset)) nw = min([os.cpu_count() // WORLD_SIZE, batch_size if batch_size > 1 else 0, workers]) # number of workers sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) loader = DataLoader if image_weights else InfiniteDataLoader # only DataLoader allows for attribute updates return loader(dataset, batch_size=batch_size, shuffle=shuffle and sampler is None, num_workers=nw, sampler=sampler, pin_memory=True, collate_fn=LoadImagesAndLabels.collate_fn4 if quad else LoadImagesAndLabels.collate_fn), dataset class InfiniteDataLoader(dataloader.DataLoader): """ Dataloader that reuses workers Uses same syntax as vanilla DataLoader """ def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) object.__setattr__(self, 'batch_sampler', _RepeatSampler(self.batch_sampler)) self.iterator = super().__iter__() def __len__(self): return len(self.batch_sampler.sampler) def __iter__(self): for i in range(len(self)): yield next(self.iterator) class _RepeatSampler: """ Sampler that repeats forever Args: sampler (Sampler) """ def __init__(self, sampler): self.sampler = sampler def __iter__(self): while True: yield from iter(self.sampler) class LoadImages: # YOLOv5 image/video dataloader, i.e. `python detect.py --source image.jpg/vid.mp4` def __init__(self, path, img_size=640, stride=32, auto=True): p = str(Path(path).resolve()) # os-agnostic absolute path if '*' in p: files = sorted(glob.glob(p, recursive=True)) # glob elif os.path.isdir(p): files = sorted(glob.glob(os.path.join(p, '*.*'))) # dir elif os.path.isfile(p): files = [p] # files else: raise Exception(f'ERROR: {p} does not exist') images = [x for x in files if x.split('.')[-1].lower() in IMG_FORMATS] videos = [x for x in files if x.split('.')[-1].lower() in VID_FORMATS] ni, nv = len(images), len(videos) self.img_size = img_size self.stride = stride self.files = images + videos self.nf = ni + nv # number of files self.video_flag = [False] * ni + [True] * nv self.mode = 'image' self.auto = auto if any(videos): self.new_video(videos[0]) # new video else: self.cap = None assert self.nf > 0, f'No images or videos found in {p}. ' \ f'Supported formats are:\nimages: {IMG_FORMATS}\nvideos: {VID_FORMATS}' def __iter__(self): self.count = 0 return self def __next__(self): if self.count == self.nf: raise StopIteration path = self.files[self.count] if self.video_flag[self.count]: # Read video self.mode = 'video' ret_val, img0 = self.cap.read() while not ret_val: self.count += 1 self.cap.release() if self.count == self.nf: # last video raise StopIteration else: path = self.files[self.count] self.new_video(path) ret_val, img0 = self.cap.read() self.frame += 1 s = f'video {self.count + 1}/{self.nf} ({self.frame}/{self.frames}) {path}: ' else: # Read image self.count += 1 img0 = cv2.imread(path) # BGR assert img0 is not None, f'Image Not Found {path}' s = f'image {self.count}/{self.nf} {path}: ' # Padded resize img = letterbox(img0, self.img_size, stride=self.stride, auto=self.auto)[0] # Convert img = img.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB img = np.ascontiguousarray(img) return path, img, img0, self.cap, s def new_video(self, path): self.frame = 0 self.cap = cv2.VideoCapture(path) self.frames = int(self.cap.get(cv2.CAP_PROP_FRAME_COUNT)) def __len__(self): return self.nf # number of files class LoadWebcam: # for inference # YOLOv5 local webcam dataloader, i.e. `python detect.py --source 0` def __init__(self, pipe='0', img_size=640, stride=32): self.img_size = img_size self.stride = stride self.pipe = eval(pipe) if pipe.isnumeric() else pipe self.cap = cv2.VideoCapture(self.pipe) # video capture object self.cap.set(cv2.CAP_PROP_BUFFERSIZE, 3) # set buffer size def __iter__(self): self.count = -1 return self def __next__(self): self.count += 1 if cv2.waitKey(1) == ord('q'): # q to quit self.cap.release() cv2.destroyAllWindows() raise StopIteration # Read frame ret_val, img0 = self.cap.read() img0 = cv2.flip(img0, 1) # flip left-right # Print assert ret_val, f'Camera Error {self.pipe}' img_path = 'webcam.jpg' s = f'webcam {self.count}: ' # Padded resize img = letterbox(img0, self.img_size, stride=self.stride)[0] # Convert img = img.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB img = np.ascontiguousarray(img) return img_path, img, img0, None, s def __len__(self): return 0 class LoadStreams: # YOLOv5 streamloader, i.e. `python detect.py --source 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP streams` def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True): self.mode = 'stream' self.img_size = img_size self.stride = stride if os.path.isfile(sources): with open(sources) as f: sources = [x.strip() for x in f.read().strip().splitlines() if len(x.strip())] else: sources = [sources] n = len(sources) self.imgs, self.fps, self.frames, self.threads = [None] * n, [0] * n, [0] * n, [None] * n self.sources = [clean_str(x) for x in sources] # clean source names for later self.auto = auto for i, s in enumerate(sources): # index, source # Start thread to read frames from video stream st = f'{i + 1}/{n}: {s}... ' if 'youtube.com/' in s or 'youtu.be/' in s: # if source is YouTube video check_requirements(('pafy', 'youtube_dl')) import pafy s = pafy.new(s).getbest(preftype="mp4").url # YouTube URL s = eval(s) if s.isnumeric() else s # i.e. s = '0' local webcam cap = cv2.VideoCapture(s) assert cap.isOpened(), f'{st}Failed to open {s}' w = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)) h = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) self.fps[i] = max(cap.get(cv2.CAP_PROP_FPS) % 100, 0) or 30.0 # 30 FPS fallback self.frames[i] = max(int(cap.get(cv2.CAP_PROP_FRAME_COUNT)), 0) or float('inf') # infinite stream fallback _, self.imgs[i] = cap.read() # guarantee first frame self.threads[i] = Thread(target=self.update, args=([i, cap, s]), daemon=True) LOGGER.info(f"{st} Success ({self.frames[i]} frames {w}x{h} at {self.fps[i]:.2f} FPS)") self.threads[i].start() LOGGER.info('') # newline # check for common shapes s = np.stack([letterbox(x, self.img_size, stride=self.stride, auto=self.auto)[0].shape for x in self.imgs]) self.rect = np.unique(s, axis=0).shape[0] == 1 # rect inference if all shapes equal if not self.rect: LOGGER.warning('WARNING: Stream shapes differ. For optimal performance supply similarly-shaped streams.') def update(self, i, cap, stream): # Read stream `i` frames in daemon thread n, f, read = 0, self.frames[i], 1 # frame number, frame array, inference every 'read' frame while cap.isOpened() and n < f: n += 1 # _, self.imgs[index] = cap.read() cap.grab() if n % read == 0: success, im = cap.retrieve() if success: self.imgs[i] = im else: LOGGER.warning('WARNING: Video stream unresponsive, please check your IP camera connection.') self.imgs[i] = np.zeros_like(self.imgs[i]) cap.open(stream) # re-open stream if signal was lost time.sleep(1 / self.fps[i]) # wait time def __iter__(self): self.count = -1 return self def __next__(self): self.count += 1 if not all(x.is_alive() for x in self.threads) or cv2.waitKey(1) == ord('q'): # q to quit cv2.destroyAllWindows() raise StopIteration # Letterbox img0 = self.imgs.copy() img = [letterbox(x, self.img_size, stride=self.stride, auto=self.rect and self.auto)[0] for x in img0] # Stack img = np.stack(img, 0) # Convert img = img[..., ::-1].transpose((0, 3, 1, 2)) # BGR to RGB, BHWC to BCHW img = np.ascontiguousarray(img) return self.sources, img, img0, None, '' def __len__(self): return len(self.sources) # 1E12 frames = 32 streams at 30 FPS for 30 years def img2label_paths(img_paths): # Define label paths as a function of image paths sa, sb = os.sep + 'images' + os.sep, os.sep + 'labels' + os.sep # /images/, /labels/ substrings return [sb.join(x.rsplit(sa, 1)).rsplit('.', 1)[0] + '.txt' for x in img_paths] class LoadImagesAndLabels(Dataset): # YOLOv5 train_loader/val_loader, loads images and labels for training and validation cache_version = 0.6 # dataset labels *.cache version def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, rect=False, image_weights=False, cache_images=False, single_cls=False, stride=32, pad=0.0, prefix=''): self.img_size = img_size self.augment = augment self.hyp = hyp self.image_weights = image_weights self.rect = False if image_weights else rect self.mosaic = self.augment and not self.rect # load 4 images at a time into a mosaic (only during training) self.mosaic_border = [-img_size // 2, -img_size // 2] self.stride = stride self.path = path self.albumentations = Albumentations() if augment else None try: f = [] # image files for p in path if isinstance(path, list) else [path]: p = Path(p) # os-agnostic if p.is_dir(): # dir f += glob.glob(str(p / '**' / '*.*'), recursive=True) # f = list(p.rglob('*.*')) # pathlib elif p.is_file(): # file with open(p) as t: t = t.read().strip().splitlines() parent = str(p.parent) + os.sep f += [x.replace('./', parent) if x.startswith('./') else x for x in t] # local to global path # f += [p.parent / x.lstrip(os.sep) for x in t] # local to global path (pathlib) else: raise Exception(f'{prefix}{p} does not exist') self.img_files = sorted(x.replace('/', os.sep) for x in f if x.split('.')[-1].lower() in IMG_FORMATS) # self.img_files = sorted([x for x in f if x.suffix[1:].lower() in IMG_FORMATS]) # pathlib assert self.img_files, f'{prefix}No images found' except Exception as e: raise Exception(f'{prefix}Error loading data from {path}: {e}\nSee {HELP_URL}') # Check cache self.label_files = img2label_paths(self.img_files) # labels cache_path = (p if p.is_file() else Path(self.label_files[0]).parent).with_suffix('.cache') try: cache, exists = np.load(cache_path, allow_pickle=True).item(), True # load dict assert cache['version'] == self.cache_version # same version assert cache['hash'] == get_hash(self.label_files + self.img_files) # same hash except: cache, exists = self.cache_labels(cache_path, prefix), False # cache # Display cache nf, nm, ne, nc, n = cache.pop('results') # found, missing, empty, corrupted, total if exists: d = f"Scanning '{cache_path}' images and labels... {nf} found, {nm} missing, {ne} empty, {nc} corrupted" tqdm(None, desc=prefix + d, total=n, initial=n) # display cache results if cache['msgs']: LOGGER.info('\n'.join(cache['msgs'])) # display warnings assert nf > 0 or not augment, f'{prefix}No labels in {cache_path}. Can not train without labels. See {HELP_URL}' # Read cache [cache.pop(k) for k in ('hash', 'version', 'msgs')] # remove items labels, shapes, self.segments = zip(*cache.values()) self.labels = list(labels) self.shapes = np.array(shapes, dtype=np.float64) self.img_files = list(cache.keys()) # update self.label_files = img2label_paths(cache.keys()) # update n = len(shapes) # number of images bi = np.floor(np.arange(n) / batch_size).astype(np.int) # batch index nb = bi[-1] + 1 # number of batches self.batch = bi # batch index of image self.n = n self.indices = range(n) # Update labels include_class = [] # filter labels to include only these classes (optional) include_class_array = np.array(include_class).reshape(1, -1) for i, (label, segment) in enumerate(zip(self.labels, self.segments)): if include_class: j = (label[:, 0:1] == include_class_array).any(1) self.labels[i] = label[j] if segment: self.segments[i] = segment[j] if single_cls: # single-class training, merge all classes into 0 self.labels[i][:, 0] = 0 if segment: self.segments[i][:, 0] = 0 # Rectangular Training if self.rect: # Sort by aspect ratio s = self.shapes # wh ar = s[:, 1] / s[:, 0] # aspect ratio irect = ar.argsort() self.img_files = [self.img_files[i] for i in irect] self.label_files = [self.label_files[i] for i in irect] self.labels = [self.labels[i] for i in irect] self.shapes = s[irect] # wh ar = ar[irect] # Set training image shapes shapes = [[1, 1]] * nb for i in range(nb): ari = ar[bi == i] mini, maxi = ari.min(), ari.max() if maxi < 1: shapes[i] = [maxi, 1] elif mini > 1: shapes[i] = [1, 1 / mini] self.batch_shapes = np.ceil(np.array(shapes) * img_size / stride + pad).astype(np.int) * stride # Cache images into memory for faster training (WARNING: large datasets may exceed system RAM) self.imgs, self.img_npy = [None] * n, [None] * n if cache_images: if cache_images == 'disk': self.im_cache_dir = Path(Path(self.img_files[0]).parent.as_posix() + '_npy') self.img_npy = [self.im_cache_dir / Path(f).with_suffix('.npy').name for f in self.img_files] self.im_cache_dir.mkdir(parents=True, exist_ok=True) gb = 0 # Gigabytes of cached images self.img_hw0, self.img_hw = [None] * n, [None] * n results = ThreadPool(NUM_THREADS).imap(lambda x: load_image(*x), zip(repeat(self), range(n))) pbar = tqdm(enumerate(results), total=n) for i, x in pbar: if cache_images == 'disk': if not self.img_npy[i].exists(): np.save(self.img_npy[i].as_posix(), x[0]) gb += self.img_npy[i].stat().st_size else: self.imgs[i], self.img_hw0[i], self.img_hw[i] = x # im, hw_orig, hw_resized = load_image(self, i) gb += self.imgs[i].nbytes pbar.desc = f'{prefix}Caching images ({gb / 1E9:.1f}GB {cache_images})' pbar.close() def cache_labels(self, path=Path('./labels.cache'), prefix=''): # Cache dataset labels, check images and read shapes x = {} # dict nm, nf, ne, nc, msgs = 0, 0, 0, 0, [] # number missing, found, empty, corrupt, messages desc = f"{prefix}Scanning '{path.parent / path.stem}' images and labels..." with Pool(NUM_THREADS) as pool: pbar = tqdm(pool.imap(verify_image_label, zip(self.img_files, self.label_files, repeat(prefix))), desc=desc, total=len(self.img_files)) for im_file, l, shape, segments, nm_f, nf_f, ne_f, nc_f, msg in pbar: nm += nm_f nf += nf_f ne += ne_f nc += nc_f if im_file: x[im_file] = [l, shape, segments] if msg: msgs.append(msg) pbar.desc = f"{desc}{nf} found, {nm} missing, {ne} empty, {nc} corrupted" pbar.close() if msgs: LOGGER.info('\n'.join(msgs)) if nf == 0: LOGGER.warning(f'{prefix}WARNING: No labels found in {path}. See {HELP_URL}') x['hash'] = get_hash(self.label_files + self.img_files) x['results'] = nf, nm, ne, nc, len(self.img_files) x['msgs'] = msgs # warnings x['version'] = self.cache_version # cache version try: np.save(path, x) # save cache for next time path.with_suffix('.cache.npy').rename(path) # remove .npy suffix LOGGER.info(f'{prefix}New cache created: {path}') except Exception as e: LOGGER.warning(f'{prefix}WARNING: Cache directory {path.parent} is not writeable: {e}') # not writeable return x def __len__(self): return len(self.img_files) # def __iter__(self): # self.count = -1 # print('ran dataset iter') # #self.shuffled_vector = np.random.permutation(self.nF) if self.augment else np.arange(self.nF) # return self def __getitem__(self, index): index = self.indices[index] # linear, shuffled, or image_weights hyp = self.hyp mosaic = self.mosaic and random.random() < hyp['mosaic'] if mosaic: # Load mosaic img, labels = load_mosaic(self, index) shapes = None # MixUp augmentation if random.random() < hyp['mixup']: img, labels = mixup(img, labels, *load_mosaic(self, random.randint(0, self.n - 1))) else: # Load image img, (h0, w0), (h, w) = load_image(self, index) # Letterbox shape = self.batch_shapes[self.batch[index]] if self.rect else self.img_size # final letterboxed shape img, ratio, pad = letterbox(img, shape, auto=False, scaleup=self.augment) shapes = (h0, w0), ((h / h0, w / w0), pad) # for COCO mAP rescaling labels = self.labels[index].copy() if labels.size: # normalized xywh to pixel xyxy format labels[:, 1:] = xywhn2xyxy(labels[:, 1:], ratio[0] * w, ratio[1] * h, padw=pad[0], padh=pad[1]) if self.augment: img, labels = random_perspective(img, labels, degrees=hyp['degrees'], translate=hyp['translate'], scale=hyp['scale'], shear=hyp['shear'], perspective=hyp['perspective']) nl = len(labels) # number of labels if nl: labels[:, 1:5] = xyxy2xywhn(labels[:, 1:5], w=img.shape[1], h=img.shape[0], clip=True, eps=1E-3) if self.augment: # Albumentations img, labels = self.albumentations(img, labels) nl = len(labels) # update after albumentations # HSV color-space augment_hsv(img, hgain=hyp['hsv_h'], sgain=hyp['hsv_s'], vgain=hyp['hsv_v']) # Flip up-down if random.random() < hyp['flipud']: img = np.flipud(img) if nl: labels[:, 2] = 1 - labels[:, 2] # Flip left-right if random.random() < hyp['fliplr']: img = np.fliplr(img) if nl: labels[:, 1] = 1 - labels[:, 1] # Cutouts # labels = cutout(img, labels, p=0.5) # nl = len(labels) # update after cutout labels_out = torch.zeros((nl, 6)) if nl: labels_out[:, 1:] = torch.from_numpy(labels) # Convert img = img.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB img = np.ascontiguousarray(img) return torch.from_numpy(img), labels_out, self.img_files[index], shapes @staticmethod def collate_fn(batch): img, label, path, shapes = zip(*batch) # transposed for i, l in enumerate(label): l[:, 0] = i # add target image index for build_targets() return torch.stack(img, 0), torch.cat(label, 0), path, shapes @staticmethod def collate_fn4(batch): img, label, path, shapes = zip(*batch) # transposed n = len(shapes) // 4 img4, label4, path4, shapes4 = [], [], path[:n], shapes[:n] ho = torch.tensor([[0.0, 0, 0, 1, 0, 0]]) wo = torch.tensor([[0.0, 0, 1, 0, 0, 0]]) s = torch.tensor([[1, 1, 0.5, 0.5, 0.5, 0.5]]) # scale for i in range(n): # zidane torch.zeros(16,3,720,1280) # BCHW i *= 4 if random.random() < 0.5: im = F.interpolate(img[i].unsqueeze(0).float(), scale_factor=2.0, mode='bilinear', align_corners=False)[ 0].type(img[i].type()) l = label[i] else: im = torch.cat((torch.cat((img[i], img[i + 1]), 1), torch.cat((img[i + 2], img[i + 3]), 1)), 2) l = torch.cat((label[i], label[i + 1] + ho, label[i + 2] + wo, label[i + 3] + ho + wo), 0) * s img4.append(im) label4.append(l) for i, l in enumerate(label4): l[:, 0] = i # add target image index for build_targets() return torch.stack(img4, 0), torch.cat(label4, 0), path4, shapes4 # Ancillary functions -------------------------------------------------------------------------------------------------- def load_image(self, i): # loads 1 image from dataset index 'i', returns im, original hw, resized hw im = self.imgs[i] if im is None: # not cached in ram npy = self.img_npy[i] if npy and npy.exists(): # load npy im = np.load(npy) else: # read image path = self.img_files[i] im = cv2.imread(path) # BGR assert im is not None, f'Image Not Found {path}' h0, w0 = im.shape[:2] # orig hw r = self.img_size / max(h0, w0) # ratio if r != 1: # if sizes are not equal im = cv2.resize(im, (int(w0 * r), int(h0 * r)), interpolation=cv2.INTER_AREA if r < 1 and not self.augment else cv2.INTER_LINEAR) return im, (h0, w0), im.shape[:2] # im, hw_original, hw_resized else: return self.imgs[i], self.img_hw0[i], self.img_hw[i] # im, hw_original, hw_resized def load_mosaic(self, index): # YOLOv5 4-mosaic loader. Loads 1 image + 3 random images into a 4-image mosaic labels4, segments4 = [], [] s = self.img_size yc, xc = (int(random.uniform(-x, 2 * s + x)) for x in self.mosaic_border) # mosaic center x, y indices = [index] + random.choices(self.indices, k=3) # 3 additional image indices random.shuffle(indices) for i, index in enumerate(indices): # Load image img, _, (h, w) = load_image(self, index) # place img in img4 if i == 0: # top left img4 = np.full((s * 2, s * 2, img.shape[2]), 114, dtype=np.uint8) # base image with 4 tiles x1a, y1a, x2a, y2a = max(xc - w, 0), max(yc - h, 0), xc, yc # xmin, ymin, xmax, ymax (large image) x1b, y1b, x2b, y2b = w - (x2a - x1a), h - (y2a - y1a), w, h # xmin, ymin, xmax, ymax (small image) elif i == 1: # top right x1a, y1a, x2a, y2a = xc, max(yc - h, 0), min(xc + w, s * 2), yc x1b, y1b, x2b, y2b = 0, h - (y2a - y1a), min(w, x2a - x1a), h elif i == 2: # bottom left x1a, y1a, x2a, y2a = max(xc - w, 0), yc, xc, min(s * 2, yc + h) x1b, y1b, x2b, y2b = w - (x2a - x1a), 0, w, min(y2a - y1a, h) elif i == 3: # bottom right x1a, y1a, x2a, y2a = xc, yc, min(xc + w, s * 2), min(s * 2, yc + h) x1b, y1b, x2b, y2b = 0, 0, min(w, x2a - x1a), min(y2a - y1a, h) img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b] # img4[ymin:ymax, xmin:xmax] padw = x1a - x1b padh = y1a - y1b # Labels labels, segments = self.labels[index].copy(), self.segments[index].copy() if labels.size: labels[:, 1:] = xywhn2xyxy(labels[:, 1:], w, h, padw, padh) # normalized xywh to pixel xyxy format segments = [xyn2xy(x, w, h, padw, padh) for x in segments] labels4.append(labels) segments4.extend(segments) # Concat/clip labels labels4 = np.concatenate(labels4, 0) for x in (labels4[:, 1:], *segments4): np.clip(x, 0, 2 * s, out=x) # clip when using random_perspective() # img4, labels4 = replicate(img4, labels4) # replicate # Augment img4, labels4, segments4 = copy_paste(img4, labels4, segments4, p=self.hyp['copy_paste']) img4, labels4 = random_perspective(img4, labels4, segments4, degrees=self.hyp['degrees'], translate=self.hyp['translate'], scale=self.hyp['scale'], shear=self.hyp['shear'], perspective=self.hyp['perspective'], border=self.mosaic_border) # border to remove return img4, labels4 def load_mosaic9(self, index): # YOLOv5 9-mosaic loader. Loads 1 image + 8 random images into a 9-image mosaic labels9, segments9 = [], [] s = self.img_size indices = [index] + random.choices(self.indices, k=8) # 8 additional image indices random.shuffle(indices) for i, index in enumerate(indices): # Load image img, _, (h, w) = load_image(self, index) # place img in img9 if i == 0: # center img9 = np.full((s * 3, s * 3, img.shape[2]), 114, dtype=np.uint8) # base image with 4 tiles h0, w0 = h, w c = s, s, s + w, s + h # xmin, ymin, xmax, ymax (base) coordinates elif i == 1: # top c = s, s - h, s + w, s elif i == 2: # top right c = s + wp, s - h, s + wp + w, s elif i == 3: # right c = s + w0, s, s + w0 + w, s + h elif i == 4: # bottom right c = s + w0, s + hp, s + w0 + w, s + hp + h elif i == 5: # bottom c = s + w0 - w, s + h0, s + w0, s + h0 + h elif i == 6: # bottom left c = s + w0 - wp - w, s + h0, s + w0 - wp, s + h0 + h elif i == 7: # left c = s - w, s + h0 - h, s, s + h0 elif i == 8: # top left c = s - w, s + h0 - hp - h, s, s + h0 - hp padx, pady = c[:2] x1, y1, x2, y2 = (max(x, 0) for x in c) # allocate coords # Labels labels, segments = self.labels[index].copy(), self.segments[index].copy() if labels.size: labels[:, 1:] = xywhn2xyxy(labels[:, 1:], w, h, padx, pady) # normalized xywh to pixel xyxy format segments = [xyn2xy(x, w, h, padx, pady) for x in segments] labels9.append(labels) segments9.extend(segments) # Image img9[y1:y2, x1:x2] = img[y1 - pady:, x1 - padx:] # img9[ymin:ymax, xmin:xmax] hp, wp = h, w # height, width previous # Offset yc, xc = (int(random.uniform(0, s)) for _ in self.mosaic_border) # mosaic center x, y img9 = img9[yc:yc + 2 * s, xc:xc + 2 * s] # Concat/clip labels labels9 = np.concatenate(labels9, 0) labels9[:, [1, 3]] -= xc labels9[:, [2, 4]] -= yc c = np.array([xc, yc]) # centers segments9 = [x - c for x in segments9] for x in (labels9[:, 1:], *segments9): np.clip(x, 0, 2 * s, out=x) # clip when using random_perspective() # img9, labels9 = replicate(img9, labels9) # replicate # Augment img9, labels9 = random_perspective(img9, labels9, segments9, degrees=self.hyp['degrees'], translate=self.hyp['translate'], scale=self.hyp['scale'], shear=self.hyp['shear'], perspective=self.hyp['perspective'], border=self.mosaic_border) # border to remove return img9, labels9 def create_folder(path='./new'): # Create folder if os.path.exists(path): shutil.rmtree(path) # delete output folder os.makedirs(path) # make new output folder def flatten_recursive(path='../datasets/coco128'): # Flatten a recursive directory by bringing all files to top level new_path = Path(path + '_flat') create_folder(new_path) for file in tqdm(glob.glob(str(Path(path)) + '/**/*.*', recursive=True)): shutil.copyfile(file, new_path / Path(file).name) def extract_boxes(path='../datasets/coco128'): # from utils.datasets import *; extract_boxes() # Convert detection dataset into classification dataset, with one directory per class path = Path(path) # images dir shutil.rmtree(path / 'classifier') if (path / 'classifier').is_dir() else None # remove existing files = list(path.rglob('*.*')) n = len(files) # number of files for im_file in tqdm(files, total=n): if im_file.suffix[1:] in IMG_FORMATS: # image im = cv2.imread(str(im_file))[..., ::-1] # BGR to RGB h, w = im.shape[:2] # labels lb_file = Path(img2label_paths([str(im_file)])[0]) if Path(lb_file).exists(): with open(lb_file) as f: lb = np.array([x.split() for x in f.read().strip().splitlines()], dtype=np.float32) # labels for j, x in enumerate(lb): c = int(x[0]) # class f = (path / 'classifier') / f'{c}' / f'{path.stem}_{im_file.stem}_{j}.jpg' # new filename if not f.parent.is_dir(): f.parent.mkdir(parents=True) b = x[1:] * [w, h, w, h] # box # b[2:] = b[2:].max() # rectangle to square b[2:] = b[2:] * 1.2 + 3 # pad b = xywh2xyxy(b.reshape(-1, 4)).ravel().astype(np.int) b[[0, 2]] = np.clip(b[[0, 2]], 0, w) # clip boxes outside of image b[[1, 3]] = np.clip(b[[1, 3]], 0, h) assert cv2.imwrite(str(f), im[b[1]:b[3], b[0]:b[2]]), f'box failure in {f}' def autosplit(path='../datasets/coco128/images', weights=(0.9, 0.1, 0.0), annotated_only=False): """ Autosplit a dataset into train/val/test splits and save path/autosplit_*.txt files Usage: from utils.datasets import *; autosplit() Arguments path: Path to images directory weights: Train, val, test weights (list, tuple) annotated_only: Only use images with an annotated txt file """ path = Path(path) # images dir files = sorted(x for x in path.rglob('*.*') if x.suffix[1:].lower() in IMG_FORMATS) # image files only n = len(files) # number of files random.seed(0) # for reproducibility indices = random.choices([0, 1, 2], weights=weights, k=n) # assign each image to a split txt = ['autosplit_train.txt', 'autosplit_val.txt', 'autosplit_test.txt'] # 3 txt files [(path.parent / x).unlink(missing_ok=True) for x in txt] # remove existing print(f'Autosplitting images from {path}' + ', using *.txt labeled images only' * annotated_only) for i, img in tqdm(zip(indices, files), total=n): if not annotated_only or Path(img2label_paths([str(img)])[0]).exists(): # check label with open(path.parent / txt[i], 'a') as f: f.write('./' + img.relative_to(path.parent).as_posix() + '\n') # add image to txt file def verify_image_label(args): # Verify one image-label pair im_file, lb_file, prefix = args nm, nf, ne, nc, msg, segments = 0, 0, 0, 0, '', [] # number (missing, found, empty, corrupt), message, segments try: # verify images im = Image.open(im_file) im.verify() # PIL verify shape = exif_size(im) # image size assert (shape[0] > 9) & (shape[1] > 9), f'image size {shape} <10 pixels' assert im.format.lower() in IMG_FORMATS, f'invalid image format {im.format}' if im.format.lower() in ('jpg', 'jpeg'): with open(im_file, 'rb') as f: f.seek(-2, 2) if f.read() != b'\xff\xd9': # corrupt JPEG ImageOps.exif_transpose(Image.open(im_file)).save(im_file, 'JPEG', subsampling=0, quality=100) msg = f'{prefix}WARNING: {im_file}: corrupt JPEG restored and saved' # verify labels if os.path.isfile(lb_file): nf = 1 # label found with open(lb_file) as f: l = [x.split() for x in f.read().strip().splitlines() if len(x)] if any([len(x) > 8 for x in l]): # is segment classes = np.array([x[0] for x in l], dtype=np.float32) segments = [np.array(x[1:], dtype=np.float32).reshape(-1, 2) for x in l] # (cls, xy1...) l = np.concatenate((classes.reshape(-1, 1), segments2boxes(segments)), 1) # (cls, xywh) l = np.array(l, dtype=np.float32) nl = len(l) if nl: assert l.shape[1] == 5, f'labels require 5 columns, {l.shape[1]} columns detected' assert (l >= 0).all(), f'negative label values {l[l < 0]}' assert (l[:, 1:] <= 1).all(), f'non-normalized or out of bounds coordinates {l[:, 1:][l[:, 1:] > 1]}' _, i = np.unique(l, axis=0, return_index=True) if len(i) < nl: # duplicate row check l = l[i] # remove duplicates if segments: segments = segments[i] msg = f'{prefix}WARNING: {im_file}: {nl - len(i)} duplicate labels removed' else: ne = 1 # label empty l = np.zeros((0, 5), dtype=np.float32) else: nm = 1 # label missing l = np.zeros((0, 5), dtype=np.float32) return im_file, l, shape, segments, nm, nf, ne, nc, msg except Exception as e: nc = 1 msg = f'{prefix}WARNING: {im_file}: ignoring corrupt image/label: {e}' return [None, None, None, None, nm, nf, ne, nc, msg] def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False, profile=False, hub=False): """ Return dataset statistics dictionary with images and instances counts per split per class To run in parent directory: export PYTHONPATH="$PWD/yolov5" Usage1: from utils.datasets import *; dataset_stats('coco128.yaml', autodownload=True) Usage2: from utils.datasets import *; dataset_stats('../datasets/coco128_with_yaml.zip') Arguments path: Path to data.yaml or data.zip (with data.yaml inside data.zip) autodownload: Attempt to download dataset if not found locally verbose: Print stats dictionary """ def round_labels(labels): # Update labels to integer class and 6 decimal place floats return [[int(c), *(round(x, 4) for x in points)] for c, *points in labels] def unzip(path): # Unzip data.zip TODO: CONSTRAINT: path/to/abc.zip MUST unzip to 'path/to/abc/' if str(path).endswith('.zip'): # path is data.zip assert Path(path).is_file(), f'Error unzipping {path}, file not found' ZipFile(path).extractall(path=path.parent) # unzip dir = path.with_suffix('') # dataset directory == zip name return True, str(dir), next(dir.rglob('*.yaml')) # zipped, data_dir, yaml_path else: # path is data.yaml return False, None, path def hub_ops(f, max_dim=1920): # HUB ops for 1 image 'f': resize and save at reduced quality in /dataset-hub for web/app viewing f_new = im_dir / Path(f).name # dataset-hub image filename try: # use PIL im = Image.open(f) r = max_dim / max(im.height, im.width) # ratio if r < 1.0: # image too large im = im.resize((int(im.width * r), int(im.height * r))) im.save(f_new, 'JPEG', quality=75, optimize=True) # save except Exception as e: # use OpenCV print(f'WARNING: HUB ops PIL failure {f}: {e}') im = cv2.imread(f) im_height, im_width = im.shape[:2] r = max_dim / max(im_height, im_width) # ratio if r < 1.0: # image too large im = cv2.resize(im, (int(im_width * r), int(im_height * r)), interpolation=cv2.INTER_AREA) cv2.imwrite(str(f_new), im) zipped, data_dir, yaml_path = unzip(Path(path)) with open(check_yaml(yaml_path), errors='ignore') as f: data = yaml.safe_load(f) # data dict if zipped: data['path'] = data_dir # TODO: should this be dir.resolve()? check_dataset(data, autodownload) # download dataset if missing hub_dir = Path(data['path'] + ('-hub' if hub else '')) stats = {'nc': data['nc'], 'names': data['names']} # statistics dictionary for split in 'train', 'val', 'test': if data.get(split) is None: stats[split] = None # i.e. no test set continue x = [] dataset = LoadImagesAndLabels(data[split]) # load dataset for label in tqdm(dataset.labels, total=dataset.n, desc='Statistics'): x.append(np.bincount(label[:, 0].astype(int), minlength=data['nc'])) x = np.array(x) # shape(128x80) stats[split] = {'instance_stats': {'total': int(x.sum()), 'per_class': x.sum(0).tolist()}, 'image_stats': {'total': dataset.n, 'unlabelled': int(np.all(x == 0, 1).sum()), 'per_class': (x > 0).sum(0).tolist()}, 'labels': [{str(Path(k).name): round_labels(v.tolist())} for k, v in zip(dataset.img_files, dataset.labels)]} if hub: im_dir = hub_dir / 'images' im_dir.mkdir(parents=True, exist_ok=True) for _ in tqdm(ThreadPool(NUM_THREADS).imap(hub_ops, dataset.img_files), total=dataset.n, desc='HUB Ops'): pass # Profile stats_path = hub_dir / 'stats.json' if profile: for _ in range(1): file = stats_path.with_suffix('.npy') t1 = time.time() np.save(file, stats) t2 = time.time() x = np.load(file, allow_pickle=True) print(f'stats.npy times: {time.time() - t2:.3f}s read, {t2 - t1:.3f}s write') file = stats_path.with_suffix('.json') t1 = time.time() with open(file, 'w') as f: json.dump(stats, f) # save stats *.json t2 = time.time() with open(file) as f: x = json.load(f) # load hyps dict print(f'stats.json times: {time.time() - t2:.3f}s read, {t2 - t1:.3f}s write') # Save, print and return if hub: print(f'Saving {stats_path.resolve()}...') with open(stats_path, 'w') as f: json.dump(stats, f) # save stats.json if verbose: print(json.dumps(stats, indent=2, sort_keys=False)) return stats ================================================ FILE: utils/downloads.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Download utils """ import os import platform import subprocess import time import urllib from pathlib import Path from zipfile import ZipFile import requests import torch def gsutil_getsize(url=''): # gs://bucket/file size https://cloud.google.com/storage/docs/gsutil/commands/du s = subprocess.check_output(f'gsutil du {url}', shell=True).decode('utf-8') return eval(s.split(' ')[0]) if len(s) else 0 # bytes def safe_download(file, url, url2=None, min_bytes=1E0, error_msg=''): # Attempts to download file from url or url2, checks and removes incomplete downloads < min_bytes file = Path(file) assert_msg = f"Downloaded file '{file}' does not exist or size is < min_bytes={min_bytes}" try: # url1 print(f'Downloading {url} to {file}...') torch.hub.download_url_to_file(url, str(file)) assert file.exists() and file.stat().st_size > min_bytes, assert_msg # check except Exception as e: # url2 file.unlink(missing_ok=True) # remove partial downloads print(f'ERROR: {e}\nRe-attempting {url2 or url} to {file}...') os.system(f"curl -L '{url2 or url}' -o '{file}' --retry 3 -C -") # curl download, retry and resume on fail finally: if not file.exists() or file.stat().st_size < min_bytes: # check file.unlink(missing_ok=True) # remove partial downloads print(f"ERROR: {assert_msg}\n{error_msg}") print('') def attempt_download(file, repo='ultralytics/yolov5'): # from utils.downloads import *; attempt_download() # Attempt file download if does not exist file = Path(str(file).strip().replace("'", '')) if not file.exists(): # URL specified name = Path(urllib.parse.unquote(str(file))).name # decode '%2F' to '/' etc. if str(file).startswith(('http:/', 'https:/')): # download url = str(file).replace(':/', '://') # Pathlib turns :// -> :/ name = name.split('?')[0] # parse authentication https://url.com/file.txt?auth... safe_download(file=name, url=url, min_bytes=1E5) return name # GitHub assets file.parent.mkdir(parents=True, exist_ok=True) # make parent dir (if required) try: response = requests.get(f'https://api.github.com/repos/{repo}/releases/latest').json() # github api assets = [x['name'] for x in response['assets']] # release assets, i.e. ['yolov5s.pt', 'yolov5m.pt', ...] tag = response['tag_name'] # i.e. 'v1.0' except: # fallback plan assets = ['yolov5n.pt', 'yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt', 'yolov5n6.pt', 'yolov5s6.pt', 'yolov5m6.pt', 'yolov5l6.pt', 'yolov5x6.pt'] try: tag = subprocess.check_output('git tag', shell=True, stderr=subprocess.STDOUT).decode().split()[-1] except: tag = 'v6.0' # current release if name in assets: safe_download(file, url=f'https://github.com/{repo}/releases/download/{tag}/{name}', # url2=f'https://storage.googleapis.com/{repo}/ckpt/{name}', # backup url (optional) min_bytes=1E5, error_msg=f'{file} missing, try downloading from https://github.com/{repo}/releases/') return str(file) def gdrive_download(id='16TiPfZj7htmTyhntwcZyEEAejOUxuT6m', file='tmp.zip'): # Downloads a file from Google Drive. from yolov5.utils.downloads import *; gdrive_download() t = time.time() file = Path(file) cookie = Path('cookie') # gdrive cookie print(f'Downloading https://drive.google.com/uc?export=download&id={id} as {file}... ', end='') file.unlink(missing_ok=True) # remove existing file cookie.unlink(missing_ok=True) # remove existing cookie # Attempt file download out = "NUL" if platform.system() == "Windows" else "/dev/null" os.system(f'curl -c ./cookie -s -L "drive.google.com/uc?export=download&id={id}" > {out}') if os.path.exists('cookie'): # large file s = f'curl -Lb ./cookie "drive.google.com/uc?export=download&confirm={get_token()}&id={id}" -o {file}' else: # small file s = f'curl -s -L -o {file} "drive.google.com/uc?export=download&id={id}"' r = os.system(s) # execute, capture return cookie.unlink(missing_ok=True) # remove existing cookie # Error check if r != 0: file.unlink(missing_ok=True) # remove partial print('Download error ') # raise Exception('Download error') return r # Unzip if archive if file.suffix == '.zip': print('unzipping... ', end='') ZipFile(file).extractall(path=file.parent) # unzip file.unlink() # remove zip print(f'Done ({time.time() - t:.1f}s)') return r def get_token(cookie="./cookie"): with open(cookie) as f: for line in f: if "download" in line: return line.split()[-1] return "" # Google utils: https://cloud.google.com/storage/docs/reference/libraries ---------------------------------------------- # # # def upload_blob(bucket_name, source_file_name, destination_blob_name): # # Uploads a file to a bucket # # https://cloud.google.com/storage/docs/uploading-objects#storage-upload-object-python # # storage_client = storage.Client() # bucket = storage_client.get_bucket(bucket_name) # blob = bucket.blob(destination_blob_name) # # blob.upload_from_filename(source_file_name) # # print('File {} uploaded to {}.'.format( # source_file_name, # destination_blob_name)) # # # def download_blob(bucket_name, source_blob_name, destination_file_name): # # Uploads a blob from a bucket # storage_client = storage.Client() # bucket = storage_client.get_bucket(bucket_name) # blob = bucket.blob(source_blob_name) # # blob.download_to_filename(destination_file_name) # # print('Blob {} downloaded to {}.'.format( # source_blob_name, # destination_file_name)) ================================================ FILE: utils/flask_rest_api/README.md ================================================ # Flask REST API [REST](https://en.wikipedia.org/wiki/Representational_state_transfer) [API](https://en.wikipedia.org/wiki/API)s are commonly used to expose Machine Learning (ML) models to other services. This folder contains an example REST API created using Flask to expose the YOLOv5s model from [PyTorch Hub](https://pytorch.org/hub/ultralytics_yolov5/). ## Requirements [Flask](https://palletsprojects.com/p/flask/) is required. Install with: ```shell $ pip install Flask ``` ## Run After Flask installation run: ```shell $ python3 restapi.py --port 5000 ``` Then use [curl](https://curl.se/) to perform a request: ```shell $ curl -X POST -F image=@zidane.jpg 'http://localhost:5000/v1/object-detection/yolov5s' ``` The model inference results are returned as a JSON response: ```json [ { "class": 0, "confidence": 0.8900438547, "height": 0.9318675399, "name": "person", "width": 0.3264600933, "xcenter": 0.7438579798, "ycenter": 0.5207948685 }, { "class": 0, "confidence": 0.8440024257, "height": 0.7155083418, "name": "person", "width": 0.6546785235, "xcenter": 0.427829951, "ycenter": 0.6334488392 }, { "class": 27, "confidence": 0.3771208823, "height": 0.3902671337, "name": "tie", "width": 0.0696444362, "xcenter": 0.3675483763, "ycenter": 0.7991207838 }, { "class": 27, "confidence": 0.3527112305, "height": 0.1540903747, "name": "tie", "width": 0.0336618312, "xcenter": 0.7814827561, "ycenter": 0.5065554976 } ] ``` An example python script to perform inference using [requests](https://docs.python-requests.org/en/master/) is given in `example_request.py` ================================================ FILE: utils/flask_rest_api/example_request.py ================================================ """Perform test request""" import pprint import requests DETECTION_URL = "http://localhost:5000/v1/object-detection/yolov5s" TEST_IMAGE = "zidane.jpg" image_data = open(TEST_IMAGE, "rb").read() response = requests.post(DETECTION_URL, files={"image": image_data}).json() pprint.pprint(response) ================================================ FILE: utils/flask_rest_api/restapi.py ================================================ """ Run a rest API exposing the yolov5s object detection model """ import argparse import io import torch from flask import Flask, request from PIL import Image app = Flask(__name__) DETECTION_URL = "/v1/object-detection/yolov5s" @app.route(DETECTION_URL, methods=["POST"]) def predict(): if not request.method == "POST": return if request.files.get("image"): image_file = request.files["image"] image_bytes = image_file.read() img = Image.open(io.BytesIO(image_bytes)) results = model(img, size=640) # reduce size=320 for faster inference return results.pandas().xyxy[0].to_json(orient="records") if __name__ == "__main__": parser = argparse.ArgumentParser(description="Flask API exposing YOLOv5 model") parser.add_argument("--port", default=5000, type=int, help="port number") args = parser.parse_args() model = torch.hub.load("ultralytics/yolov5", "yolov5s", force_reload=True) # force_reload to recache app.run(host="0.0.0.0", port=args.port) # debug=True causes Restarting with stat ================================================ FILE: utils/general.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ General utils """ import contextlib import glob import logging import math import os import platform import random import re import shutil import signal import time import urllib from itertools import repeat from multiprocessing.pool import ThreadPool from pathlib import Path from subprocess import check_output from zipfile import ZipFile import cv2 import numpy as np import pandas as pd import pkg_resources as pkg import torch import torchvision import yaml from utils.downloads import gsutil_getsize from utils.metrics import box_iou, fitness # Settings FILE = Path(__file__).resolve() ROOT = FILE.parents[1] # YOLOv5 root directory NUM_THREADS = min(8, max(1, os.cpu_count() - 1)) # number of YOLOv5 multiprocessing threads torch.set_printoptions(linewidth=320, precision=5, profile='long') np.set_printoptions(linewidth=320, formatter={'float_kind': '{:11.5g}'.format}) # format short g, %precision=5 pd.options.display.max_columns = 10 cv2.setNumThreads(0) # prevent OpenCV from multithreading (incompatible with PyTorch DataLoader) os.environ['NUMEXPR_MAX_THREADS'] = str(NUM_THREADS) # NumExpr max threads def set_logging(name=None, verbose=True): # Sets level and returns logger rank = int(os.getenv('RANK', -1)) # rank in world for Multi-GPU trainings logging.basicConfig(format="%(message)s", level=logging.INFO if (verbose and rank in (-1, 0)) else logging.WARNING) return logging.getLogger(name) LOGGER = set_logging(__name__) # define globally (used in train.py, val.py, detect.py, etc.) class Profile(contextlib.ContextDecorator): # Usage: @Profile() decorator or 'with Profile():' context manager def __enter__(self): self.start = time.time() def __exit__(self, type, value, traceback): print(f'Profile results: {time.time() - self.start:.5f}s') class Timeout(contextlib.ContextDecorator): # Usage: @Timeout(seconds) decorator or 'with Timeout(seconds):' context manager def __init__(self, seconds, *, timeout_msg='', suppress_timeout_errors=True): self.seconds = int(seconds) self.timeout_message = timeout_msg self.suppress = bool(suppress_timeout_errors) def _timeout_handler(self, signum, frame): raise TimeoutError(self.timeout_message) def __enter__(self): signal.signal(signal.SIGALRM, self._timeout_handler) # Set handler for SIGALRM signal.alarm(self.seconds) # start countdown for SIGALRM to be raised def __exit__(self, exc_type, exc_val, exc_tb): signal.alarm(0) # Cancel SIGALRM if it's scheduled if self.suppress and exc_type is TimeoutError: # Suppress TimeoutError return True class WorkingDirectory(contextlib.ContextDecorator): # Usage: @WorkingDirectory(dir) decorator or 'with WorkingDirectory(dir):' context manager def __init__(self, new_dir): self.dir = new_dir # new dir self.cwd = Path.cwd().resolve() # current dir def __enter__(self): os.chdir(self.dir) def __exit__(self, exc_type, exc_val, exc_tb): os.chdir(self.cwd) def try_except(func): # try-except function. Usage: @try_except decorator def handler(*args, **kwargs): try: func(*args, **kwargs) except Exception as e: print(e) return handler def methods(instance): # Get class/instance methods return [f for f in dir(instance) if callable(getattr(instance, f)) and not f.startswith("__")] def print_args(name, opt): # Print argparser arguments LOGGER.info(colorstr(f'{name}: ') + ', '.join(f'{k}={v}' for k, v in vars(opt).items())) def init_seeds(seed=0): # Initialize random number generator (RNG) seeds https://pytorch.org/docs/stable/notes/randomness.html # cudnn seed 0 settings are slower and more reproducible, else faster and less reproducible import torch.backends.cudnn as cudnn random.seed(seed) np.random.seed(seed) torch.manual_seed(seed) cudnn.benchmark, cudnn.deterministic = (False, True) if seed == 0 else (True, False) def intersect_dicts(da, db, exclude=()): # Dictionary intersection of matching keys and shapes, omitting 'exclude' keys, using da values return {k: v for k, v in da.items() if k in db and not any(x in k for x in exclude) and v.shape == db[k].shape} def get_latest_run(search_dir='.'): # Return path to most recent 'last.pt' in /runs (i.e. to --resume from) last_list = glob.glob(f'{search_dir}/**/last*.pt', recursive=True) return max(last_list, key=os.path.getctime) if last_list else '' def user_config_dir(dir='Ultralytics', env_var='YOLOV5_CONFIG_DIR'): # Return path of user configuration directory. Prefer environment variable if exists. Make dir if required. env = os.getenv(env_var) if env: path = Path(env) # use environment variable else: cfg = {'Windows': 'AppData/Roaming', 'Linux': '.config', 'Darwin': 'Library/Application Support'} # 3 OS dirs path = Path.home() / cfg.get(platform.system(), '') # OS-specific config dir path = (path if is_writeable(path) else Path('/tmp')) / dir # GCP and AWS lambda fix, only /tmp is writeable path.mkdir(exist_ok=True) # make if required return path def is_writeable(dir, test=False): # Return True if directory has write permissions, test opening a file with write permissions if test=True if test: # method 1 file = Path(dir) / 'tmp.txt' try: with open(file, 'w'): # open file with write permissions pass file.unlink() # remove file return True except OSError: return False else: # method 2 return os.access(dir, os.R_OK) # possible issues on Windows def is_docker(): # Is environment a Docker container? return Path('/workspace').exists() # or Path('/.dockerenv').exists() def is_colab(): # Is environment a Google Colab instance? try: import google.colab return True except ImportError: return False def is_pip(): # Is file in a pip package? return 'site-packages' in Path(__file__).resolve().parts def is_ascii(s=''): # Is string composed of all ASCII (no UTF) characters? (note str().isascii() introduced in python 3.7) s = str(s) # convert list, tuple, None, etc. to str return len(s.encode().decode('ascii', 'ignore')) == len(s) def is_chinese(s='人工智能'): # Is string composed of any Chinese characters? return re.search('[\u4e00-\u9fff]', s) def emojis(str=''): # Return platform-dependent emoji-safe version of string return str.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else str def file_size(path): # Return file/dir size (MB) path = Path(path) if path.is_file(): return path.stat().st_size / 1E6 elif path.is_dir(): return sum(f.stat().st_size for f in path.glob('**/*') if f.is_file()) / 1E6 else: return 0.0 def check_online(): # Check internet connectivity import socket try: socket.create_connection(("1.1.1.1", 443), 5) # check host accessibility return True except OSError: return False @try_except @WorkingDirectory(ROOT) def check_git_status(): # Recommend 'git pull' if code is out of date msg = ', for updates see https://github.com/ultralytics/yolov5' print(colorstr('github: '), end='') assert Path('.git').exists(), 'skipping check (not a git repository)' + msg assert not is_docker(), 'skipping check (Docker image)' + msg assert check_online(), 'skipping check (offline)' + msg cmd = 'git fetch && git config --get remote.origin.url' url = check_output(cmd, shell=True, timeout=5).decode().strip().rstrip('.git') # git fetch branch = check_output('git rev-parse --abbrev-ref HEAD', shell=True).decode().strip() # checked out n = int(check_output(f'git rev-list {branch}..origin/master --count', shell=True)) # commits behind if n > 0: s = f"⚠️ YOLOv5 is out of date by {n} commit{'s' * (n > 1)}. Use `git pull` or `git clone {url}` to update." else: s = f'up to date with {url} ✅' print(emojis(s)) # emoji-safe def check_python(minimum='3.6.2'): # Check current python version vs. required python version check_version(platform.python_version(), minimum, name='Python ', hard=True) def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=False, hard=False): # Check version vs. required version current, minimum = (pkg.parse_version(x) for x in (current, minimum)) result = (current == minimum) if pinned else (current >= minimum) # bool if hard: # assert min requirements met assert result, f'{name}{minimum} required by YOLOv5, but {name}{current} is currently installed' else: return result @try_except def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), install=True): # Check installed dependencies meet requirements (pass *.txt file or list of packages) prefix = colorstr('red', 'bold', 'requirements:') check_python() # check python version if isinstance(requirements, (str, Path)): # requirements.txt file file = Path(requirements) assert file.exists(), f"{prefix} {file.resolve()} not found, check failed." with file.open() as f: requirements = [f'{x.name}{x.specifier}' for x in pkg.parse_requirements(f) if x.name not in exclude] else: # list or tuple of packages requirements = [x for x in requirements if x not in exclude] n = 0 # number of packages updates for r in requirements: try: pkg.require(r) except Exception as e: # DistributionNotFound or VersionConflict if requirements not met s = f"{prefix} {r} not found and is required by YOLOv5" if install: print(f"{s}, attempting auto-update...") try: assert check_online(), f"'pip install {r}' skipped (offline)" print(check_output(f"pip install '{r}'", shell=True).decode()) n += 1 except Exception as e: print(f'{prefix} {e}') else: print(f'{s}. Please install and rerun your command.') if n: # if packages updated source = file.resolve() if 'file' in locals() else requirements s = f"{prefix} {n} package{'s' * (n > 1)} updated per {source}\n" \ f"{prefix} ⚠️ {colorstr('bold', 'Restart runtime or rerun command for updates to take effect')}\n" print(emojis(s)) def check_img_size(imgsz, s=32, floor=0): # Verify image size is a multiple of stride s in each dimension if isinstance(imgsz, int): # integer i.e. img_size=640 new_size = max(make_divisible(imgsz, int(s)), floor) else: # list i.e. img_size=[640, 480] new_size = [max(make_divisible(x, int(s)), floor) for x in imgsz] if new_size != imgsz: print(f'WARNING: --img-size {imgsz} must be multiple of max stride {s}, updating to {new_size}') return new_size def check_imshow(): # Check if environment supports image displays try: assert not is_docker(), 'cv2.imshow() is disabled in Docker environments' assert not is_colab(), 'cv2.imshow() is disabled in Google Colab environments' cv2.imshow('test', np.zeros((1, 1, 3))) cv2.waitKey(1) cv2.destroyAllWindows() cv2.waitKey(1) return True except Exception as e: print(f'WARNING: Environment does not support cv2.imshow() or PIL Image.show() image displays\n{e}') return False def check_suffix(file='yolov5s.pt', suffix=('.pt',), msg=''): # Check file(s) for acceptable suffix if file and suffix: if isinstance(suffix, str): suffix = [suffix] for f in file if isinstance(file, (list, tuple)) else [file]: s = Path(f).suffix.lower() # file suffix if len(s): assert s in suffix, f"{msg}{f} acceptable suffix is {suffix}" def check_yaml(file, suffix=('.yaml', '.yml')): # Search/download YAML file (if necessary) and return path, checking suffix return check_file(file, suffix) def check_file(file, suffix=''): # Search/download file (if necessary) and return path check_suffix(file, suffix) # optional file = str(file) # convert to str() if Path(file).is_file() or file == '': # exists return file elif file.startswith(('http:/', 'https:/')): # download url = str(Path(file)).replace(':/', '://') # Pathlib turns :// -> :/ file = Path(urllib.parse.unquote(file).split('?')[0]).name # '%2F' to '/', split https://url.com/file.txt?auth if Path(file).is_file(): print(f'Found {url} locally at {file}') # file already exists else: print(f'Downloading {url} to {file}...') torch.hub.download_url_to_file(url, file) assert Path(file).exists() and Path(file).stat().st_size > 0, f'File download failed: {url}' # check return file else: # search files = [] for d in 'data', 'models', 'utils': # search directories files.extend(glob.glob(str(ROOT / d / '**' / file), recursive=True)) # find file assert len(files), f'File not found: {file}' # assert file was found assert len(files) == 1, f"Multiple files match '{file}', specify exact path: {files}" # assert unique return files[0] # return file def check_dataset(data, autodownload=True): # Download and/or unzip dataset if not found locally # Usage: https://github.com/ultralytics/yolov5/releases/download/v1.0/coco128_with_yaml.zip # Download (optional) extract_dir = '' if isinstance(data, (str, Path)) and str(data).endswith('.zip'): # i.e. gs://bucket/dir/coco128.zip download(data, dir='../datasets', unzip=True, delete=False, curl=False, threads=1) data = next((Path('../datasets') / Path(data).stem).rglob('*.yaml')) extract_dir, autodownload = data.parent, False # Read yaml (optional) if isinstance(data, (str, Path)): with open(data, errors='ignore') as f: data = yaml.safe_load(f) # dictionary # Parse yaml path = extract_dir or Path(data.get('path') or '') # optional 'path' default to '.' for k in 'train', 'val', 'test': if data.get(k): # prepend path data[k] = str(path / data[k]) if isinstance(data[k], str) else [str(path / x) for x in data[k]] assert 'nc' in data, "Dataset 'nc' key missing." if 'names' not in data: data['names'] = [f'class{i}' for i in range(data['nc'])] # assign class names if missing train, val, test, s = (data.get(x) for x in ('train', 'val', 'test', 'download')) if val: val = [Path(x).resolve() for x in (val if isinstance(val, list) else [val])] # val path if not all(x.exists() for x in val): print('\nWARNING: Dataset not found, nonexistent paths: %s' % [str(x) for x in val if not x.exists()]) if s and autodownload: # download script root = path.parent if 'path' in data else '..' # unzip directory i.e. '../' if s.startswith('http') and s.endswith('.zip'): # URL f = Path(s).name # filename print(f'Downloading {s} to {f}...') torch.hub.download_url_to_file(s, f) Path(root).mkdir(parents=True, exist_ok=True) # create root ZipFile(f).extractall(path=root) # unzip Path(f).unlink() # remove zip r = None # success elif s.startswith('bash '): # bash script print(f'Running {s} ...') r = os.system(s) else: # python script r = exec(s, {'yaml': data}) # return None print(f"Dataset autodownload {f'success, saved to {root}' if r in (0, None) else 'failure'}\n") else: raise Exception('Dataset not found.') return data # dictionary def url2file(url): # Convert URL to filename, i.e. https://url.com/file.txt?auth -> file.txt url = str(Path(url)).replace(':/', '://') # Pathlib turns :// -> :/ file = Path(urllib.parse.unquote(url)).name.split('?')[0] # '%2F' to '/', split https://url.com/file.txt?auth return file def download(url, dir='.', unzip=True, delete=True, curl=False, threads=1): # Multi-threaded file download and unzip function, used in data.yaml for autodownload def download_one(url, dir): # Download 1 file f = dir / Path(url).name # filename if Path(url).is_file(): # exists in current path Path(url).rename(f) # move to dir elif not f.exists(): print(f'Downloading {url} to {f}...') if curl: os.system(f"curl -L '{url}' -o '{f}' --retry 9 -C -") # curl download, retry and resume on fail else: torch.hub.download_url_to_file(url, f, progress=True) # torch download if unzip and f.suffix in ('.zip', '.gz'): print(f'Unzipping {f}...') if f.suffix == '.zip': ZipFile(f).extractall(path=dir) # unzip elif f.suffix == '.gz': os.system(f'tar xfz {f} --directory {f.parent}') # unzip if delete: f.unlink() # remove zip dir = Path(dir) dir.mkdir(parents=True, exist_ok=True) # make directory if threads > 1: pool = ThreadPool(threads) pool.imap(lambda x: download_one(*x), zip(url, repeat(dir))) # multi-threaded pool.close() pool.join() else: for u in [url] if isinstance(url, (str, Path)) else url: download_one(u, dir) def make_divisible(x, divisor): # Returns nearest x divisible by divisor if isinstance(divisor, torch.Tensor): divisor = int(divisor.max()) # to int return math.ceil(x / divisor) * divisor def clean_str(s): # Cleans a string by replacing special characters with underscore _ return re.sub(pattern="[|@#!¡·$€%&()=?¿^*;:,¨´><+]", repl="_", string=s) def one_cycle(y1=0.0, y2=1.0, steps=100): # lambda function for sinusoidal ramp from y1 to y2 https://arxiv.org/pdf/1812.01187.pdf return lambda x: ((1 - math.cos(x * math.pi / steps)) / 2) * (y2 - y1) + y1 def colorstr(*input): # Colors a string https://en.wikipedia.org/wiki/ANSI_escape_code, i.e. colorstr('blue', 'hello world') *args, string = input if len(input) > 1 else ('blue', 'bold', input[0]) # color arguments, string colors = {'black': '\033[30m', # basic colors 'red': '\033[31m', 'green': '\033[32m', 'yellow': '\033[33m', 'blue': '\033[34m', 'magenta': '\033[35m', 'cyan': '\033[36m', 'white': '\033[37m', 'bright_black': '\033[90m', # bright colors 'bright_red': '\033[91m', 'bright_green': '\033[92m', 'bright_yellow': '\033[93m', 'bright_blue': '\033[94m', 'bright_magenta': '\033[95m', 'bright_cyan': '\033[96m', 'bright_white': '\033[97m', 'end': '\033[0m', # misc 'bold': '\033[1m', 'underline': '\033[4m'} return ''.join(colors[x] for x in args) + f'{string}' + colors['end'] def labels_to_class_weights(labels, nc=80): # Get class weights (inverse frequency) from training labels if labels[0] is None: # no labels loaded return torch.Tensor() labels = np.concatenate(labels, 0) # labels.shape = (866643, 5) for COCO classes = labels[:, 0].astype(np.int) # labels = [class xywh] weights = np.bincount(classes, minlength=nc) # occurrences per class # Prepend gridpoint count (for uCE training) # gpi = ((320 / 32 * np.array([1, 2, 4])) ** 2 * 3).sum() # gridpoints per image # weights = np.hstack([gpi * len(labels) - weights.sum() * 9, weights * 9]) ** 0.5 # prepend gridpoints to start weights[weights == 0] = 1 # replace empty bins with 1 weights = 1 / weights # number of targets per class weights /= weights.sum() # normalize return torch.from_numpy(weights) def labels_to_image_weights(labels, nc=80, class_weights=np.ones(80)): # Produces image weights based on class_weights and image contents class_counts = np.array([np.bincount(x[:, 0].astype(np.int), minlength=nc) for x in labels]) image_weights = (class_weights.reshape(1, nc) * class_counts).sum(1) # index = random.choices(range(n), weights=image_weights, k=1) # weight image sample return image_weights def coco80_to_coco91_class(): # converts 80-index (val2014) to 91-index (paper) # https://tech.amikelive.com/node-718/what-object-categories-labels-are-in-coco-dataset/ # a = np.loadtxt('data/coco.names', dtype='str', delimiter='\n') # b = np.loadtxt('data/coco_paper.names', dtype='str', delimiter='\n') # x1 = [list(a[i] == b).index(True) + 1 for i in range(80)] # darknet to coco # x2 = [list(b[i] == a).index(True) if any(b[i] == a) else None for i in range(91)] # coco to darknet x = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 84, 85, 86, 87, 88, 89, 90] return x def xyxy2xywh(x): # Convert nx4 boxes from [x1, y1, x2, y2] to [x, y, w, h] where xy1=top-left, xy2=bottom-right y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x) y[:, 0] = (x[:, 0] + x[:, 2]) / 2 # x center y[:, 1] = (x[:, 1] + x[:, 3]) / 2 # y center y[:, 2] = x[:, 2] - x[:, 0] # width y[:, 3] = x[:, 3] - x[:, 1] # height return y def xywh2xyxy(x): # Convert nx4 boxes from [x, y, w, h] to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x) y[:, 0] = x[:, 0] - x[:, 2] / 2 # top left x y[:, 1] = x[:, 1] - x[:, 3] / 2 # top left y y[:, 2] = x[:, 0] + x[:, 2] / 2 # bottom right x y[:, 3] = x[:, 1] + x[:, 3] / 2 # bottom right y return y def xywhn2xyxy(x, w=640, h=640, padw=0, padh=0): # Convert nx4 boxes from [x, y, w, h] normalized to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x) y[:, 0] = w * (x[:, 0] - x[:, 2] / 2) + padw # top left x y[:, 1] = h * (x[:, 1] - x[:, 3] / 2) + padh # top left y y[:, 2] = w * (x[:, 0] + x[:, 2] / 2) + padw # bottom right x y[:, 3] = h * (x[:, 1] + x[:, 3] / 2) + padh # bottom right y return y def xyxy2xywhn(x, w=640, h=640, clip=False, eps=0.0): # Convert nx4 boxes from [x1, y1, x2, y2] to [x, y, w, h] normalized where xy1=top-left, xy2=bottom-right if clip: clip_coords(x, (h - eps, w - eps)) # warning: inplace clip y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x) y[:, 0] = ((x[:, 0] + x[:, 2]) / 2) / w # x center y[:, 1] = ((x[:, 1] + x[:, 3]) / 2) / h # y center y[:, 2] = (x[:, 2] - x[:, 0]) / w # width y[:, 3] = (x[:, 3] - x[:, 1]) / h # height return y def xyn2xy(x, w=640, h=640, padw=0, padh=0): # Convert normalized segments into pixel segments, shape (n,2) y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x) y[:, 0] = w * x[:, 0] + padw # top left x y[:, 1] = h * x[:, 1] + padh # top left y return y def segment2box(segment, width=640, height=640): # Convert 1 segment label to 1 box label, applying inside-image constraint, i.e. (xy1, xy2, ...) to (xyxy) x, y = segment.T # segment xy inside = (x >= 0) & (y >= 0) & (x <= width) & (y <= height) x, y, = x[inside], y[inside] return np.array([x.min(), y.min(), x.max(), y.max()]) if any(x) else np.zeros((1, 4)) # xyxy def segments2boxes(segments): # Convert segment labels to box labels, i.e. (cls, xy1, xy2, ...) to (cls, xywh) boxes = [] for s in segments: x, y = s.T # segment xy boxes.append([x.min(), y.min(), x.max(), y.max()]) # cls, xyxy return xyxy2xywh(np.array(boxes)) # cls, xywh def resample_segments(segments, n=1000): # Up-sample an (n,2) segment for i, s in enumerate(segments): x = np.linspace(0, len(s) - 1, n) xp = np.arange(len(s)) segments[i] = np.concatenate([np.interp(x, xp, s[:, i]) for i in range(2)]).reshape(2, -1).T # segment xy return segments def scale_coords(img1_shape, coords, img0_shape, ratio_pad=None): # Rescale coords (xyxy) from img1_shape to img0_shape if ratio_pad is None: # calculate from img0_shape gain = min(img1_shape[0] / img0_shape[0], img1_shape[1] / img0_shape[1]) # gain = old / new pad = (img1_shape[1] - img0_shape[1] * gain) / 2, (img1_shape[0] - img0_shape[0] * gain) / 2 # wh padding else: gain = ratio_pad[0][0] pad = ratio_pad[1] coords[:, [0, 2]] -= pad[0] # x padding coords[:, [1, 3]] -= pad[1] # y padding coords[:, :4] /= gain clip_coords(coords, img0_shape) return coords def clip_coords(boxes, shape): # Clip bounding xyxy bounding boxes to image shape (height, width) if isinstance(boxes, torch.Tensor): # faster individually boxes[:, 0].clamp_(0, shape[1]) # x1 boxes[:, 1].clamp_(0, shape[0]) # y1 boxes[:, 2].clamp_(0, shape[1]) # x2 boxes[:, 3].clamp_(0, shape[0]) # y2 else: # np.array (faster grouped) boxes[:, [0, 2]] = boxes[:, [0, 2]].clip(0, shape[1]) # x1, x2 boxes[:, [1, 3]] = boxes[:, [1, 3]].clip(0, shape[0]) # y1, y2 def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=None, agnostic=False, multi_label=False, labels=(), max_det=300): """Runs Non-Maximum Suppression (NMS) on inference results Returns: list of detections, on (n,6) tensor per image [xyxy, conf, cls] """ nc = prediction.shape[2] - 5 # number of classes xc = prediction[..., 4] > conf_thres # candidates # Checks assert 0 <= conf_thres <= 1, f'Invalid Confidence threshold {conf_thres}, valid values are between 0.0 and 1.0' assert 0 <= iou_thres <= 1, f'Invalid IoU {iou_thres}, valid values are between 0.0 and 1.0' # Settings min_wh, max_wh = 2, 4096 # (pixels) minimum and maximum box width and height max_nms = 30000 # maximum number of boxes into torchvision.ops.nms() time_limit = 10.0 # seconds to quit after redundant = True # require redundant detections multi_label &= nc > 1 # multiple labels per box (adds 0.5ms/img) merge = False # use merge-NMS t = time.time() output = [torch.zeros((0, 6), device=prediction.device)] * prediction.shape[0] for xi, x in enumerate(prediction): # image index, image inference # Apply constraints # x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0 # width-height x = x[xc[xi]] # confidence # Cat apriori labels if autolabelling if labels and len(labels[xi]): l = labels[xi] v = torch.zeros((len(l), nc + 5), device=x.device) v[:, :4] = l[:, 1:5] # box v[:, 4] = 1.0 # conf v[range(len(l)), l[:, 0].long() + 5] = 1.0 # cls x = torch.cat((x, v), 0) # If none remain process next image if not x.shape[0]: continue # Compute conf x[:, 5:] *= x[:, 4:5] # conf = obj_conf * cls_conf # Box (center x, center y, width, height) to (x1, y1, x2, y2) box = xywh2xyxy(x[:, :4]) # Detections matrix nx6 (xyxy, conf, cls) if multi_label: i, j = (x[:, 5:] > conf_thres).nonzero(as_tuple=False).T x = torch.cat((box[i], x[i, j + 5, None], j[:, None].float()), 1) else: # best class only conf, j = x[:, 5:].max(1, keepdim=True) x = torch.cat((box, conf, j.float()), 1)[conf.view(-1) > conf_thres] # Filter by class if classes is not None: x = x[(x[:, 5:6] == torch.tensor(classes, device=x.device)).any(1)] # Apply finite constraint # if not torch.isfinite(x).all(): # x = x[torch.isfinite(x).all(1)] # Check shape n = x.shape[0] # number of boxes if not n: # no boxes continue elif n > max_nms: # excess boxes x = x[x[:, 4].argsort(descending=True)[:max_nms]] # sort by confidence # Batched NMS c = x[:, 5:6] * (0 if agnostic else max_wh) # classes boxes, scores = x[:, :4] + c, x[:, 4] # boxes (offset by class), scores i = torchvision.ops.nms(boxes, scores, iou_thres) # NMS if i.shape[0] > max_det: # limit detections i = i[:max_det] if merge and (1 < n < 3E3): # Merge NMS (boxes merged using weighted mean) # update boxes as boxes(i,4) = weights(i,n) * boxes(n,4) iou = box_iou(boxes[i], boxes) > iou_thres # iou matrix weights = iou * scores[None] # box weights x[i, :4] = torch.mm(weights, x[:, :4]).float() / weights.sum(1, keepdim=True) # merged boxes if redundant: i = i[iou.sum(1) > 1] # require redundancy output[xi] = x[i] if (time.time() - t) > time_limit: print(f'WARNING: NMS time limit {time_limit}s exceeded') break # time limit exceeded return output def strip_optimizer(f='best.pt', s=''): # from utils.general import *; strip_optimizer() # Strip optimizer from 'f' to finalize training, optionally save as 's' x = torch.load(f, map_location=torch.device('cpu')) if x.get('ema'): x['model'] = x['ema'] # replace model with ema for k in 'optimizer', 'best_fitness', 'wandb_id', 'ema', 'updates': # keys x[k] = None x['epoch'] = -1 x['model'].half() # to FP16 for p in x['model'].parameters(): p.requires_grad = False torch.save(x, s or f) mb = os.path.getsize(s or f) / 1E6 # filesize print(f"Optimizer stripped from {f},{(' saved as %s,' % s) if s else ''} {mb:.1f}MB") def print_mutation(results, hyp, save_dir, bucket): evolve_csv, results_csv, evolve_yaml = save_dir / 'evolve.csv', save_dir / 'results.csv', save_dir / 'hyp_evolve.yaml' keys = ('metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', 'val/box_loss', 'val/obj_loss', 'val/cls_loss') + tuple(hyp.keys()) # [results + hyps] keys = tuple(x.strip() for x in keys) vals = results + tuple(hyp.values()) n = len(keys) # Download (optional) if bucket: url = f'gs://{bucket}/evolve.csv' if gsutil_getsize(url) > (os.path.getsize(evolve_csv) if os.path.exists(evolve_csv) else 0): os.system(f'gsutil cp {url} {save_dir}') # download evolve.csv if larger than local # Log to evolve.csv s = '' if evolve_csv.exists() else (('%20s,' * n % keys).rstrip(',') + '\n') # add header with open(evolve_csv, 'a') as f: f.write(s + ('%20.5g,' * n % vals).rstrip(',') + '\n') # Print to screen print(colorstr('evolve: ') + ', '.join(f'{x.strip():>20s}' for x in keys)) print(colorstr('evolve: ') + ', '.join(f'{x:20.5g}' for x in vals), end='\n\n\n') # Save yaml with open(evolve_yaml, 'w') as f: data = pd.read_csv(evolve_csv) data = data.rename(columns=lambda x: x.strip()) # strip keys i = np.argmax(fitness(data.values[:, :7])) # f.write('# YOLOv5 Hyperparameter Evolution Results\n' + f'# Best generation: {i}\n' + f'# Last generation: {len(data) - 1}\n' + '# ' + ', '.join(f'{x.strip():>20s}' for x in keys[:7]) + '\n' + '# ' + ', '.join(f'{x:>20.5g}' for x in data.values[i, :7]) + '\n\n') yaml.safe_dump(hyp, f, sort_keys=False) if bucket: os.system(f'gsutil cp {evolve_csv} {evolve_yaml} gs://{bucket}') # upload def apply_classifier(x, model, img, im0): # Apply a second stage classifier to YOLO outputs # Example model = torchvision.models.__dict__['efficientnet_b0'](pretrained=True).to(device).eval() im0 = [im0] if isinstance(im0, np.ndarray) else im0 for i, d in enumerate(x): # per image if d is not None and len(d): d = d.clone() # Reshape and pad cutouts b = xyxy2xywh(d[:, :4]) # boxes b[:, 2:] = b[:, 2:].max(1)[0].unsqueeze(1) # rectangle to square b[:, 2:] = b[:, 2:] * 1.3 + 30 # pad d[:, :4] = xywh2xyxy(b).long() # Rescale boxes from img_size to im0 size scale_coords(img.shape[2:], d[:, :4], im0[i].shape) # Classes pred_cls1 = d[:, 5].long() ims = [] for j, a in enumerate(d): # per item cutout = im0[i][int(a[1]):int(a[3]), int(a[0]):int(a[2])] im = cv2.resize(cutout, (224, 224)) # BGR # cv2.imwrite('example%i.jpg' % j, cutout) im = im[:, :, ::-1].transpose(2, 0, 1) # BGR to RGB, to 3x416x416 im = np.ascontiguousarray(im, dtype=np.float32) # uint8 to float32 im /= 255 # 0 - 255 to 0.0 - 1.0 ims.append(im) pred_cls2 = model(torch.Tensor(ims).to(d.device)).argmax(1) # classifier prediction x[i] = x[i][pred_cls1 == pred_cls2] # retain matching class detections return x def increment_path(path, exist_ok=False, sep='', mkdir=False): # Increment file or directory path, i.e. runs/exp --> runs/exp{sep}2, runs/exp{sep}3, ... etc. path = Path(path) # os-agnostic if path.exists() and not exist_ok: path, suffix = (path.with_suffix(''), path.suffix) if path.is_file() else (path, '') dirs = glob.glob(f"{path}{sep}*") # similar paths matches = [re.search(rf"%s{sep}(\d+)" % path.stem, d) for d in dirs] i = [int(m.groups()[0]) for m in matches if m] # indices n = max(i) + 1 if i else 2 # increment number path = Path(f"{path}{sep}{n}{suffix}") # increment path if mkdir: path.mkdir(parents=True, exist_ok=True) # make directory return path # Variables NCOLS = 0 if is_docker() else shutil.get_terminal_size().columns # terminal window size for tqdm ================================================ FILE: utils/google_app_engine/Dockerfile ================================================ FROM gcr.io/google-appengine/python # Create a virtualenv for dependencies. This isolates these packages from # system-level packages. # Use -p python3 or -p python3.7 to select python version. Default is version 2. RUN virtualenv /env -p python3 # Setting these environment variables are the same as running # source /env/bin/activate. ENV VIRTUAL_ENV /env ENV PATH /env/bin:$PATH RUN apt-get update && apt-get install -y python-opencv # Copy the application's requirements.txt and run pip to install all # dependencies into the virtualenv. ADD requirements.txt /app/requirements.txt RUN pip install -r /app/requirements.txt # Add the application source code. ADD . /app # Run a WSGI server to serve the application. gunicorn must be declared as # a dependency in requirements.txt. CMD gunicorn -b :$PORT main:app ================================================ FILE: utils/google_app_engine/additional_requirements.txt ================================================ # add these requirements in your app on top of the existing ones pip==21.1 Flask==1.0.2 gunicorn==19.9.0 ================================================ FILE: utils/google_app_engine/app.yaml ================================================ runtime: custom env: flex service: yolov5app liveness_check: initial_delay_sec: 600 manual_scaling: instances: 1 resources: cpu: 1 memory_gb: 4 disk_size_gb: 20 ================================================ FILE: utils/loggers/__init__.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Logging utils """ import os import warnings from threading import Thread import pkg_resources as pkg import torch from torch.utils.tensorboard import SummaryWriter from utils.general import colorstr, emojis from utils.loggers.wandb.wandb_utils import WandbLogger from utils.plots import plot_images, plot_results from utils.torch_utils import de_parallel LOGGERS = ('csv', 'tb', 'wandb') # text-file, TensorBoard, Weights & Biases RANK = int(os.getenv('RANK', -1)) try: import wandb assert hasattr(wandb, '__version__') # verify package import not local dir if pkg.parse_version(wandb.__version__) >= pkg.parse_version('0.12.2') and RANK in [0, -1]: try: wandb_login_success = wandb.login(timeout=30) except wandb.errors.UsageError: # known non-TTY terminal issue wandb_login_success = False if not wandb_login_success: wandb = None except (ImportError, AssertionError): wandb = None class Loggers(): # YOLOv5 Loggers class def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, include=LOGGERS): self.save_dir = save_dir self.weights = weights self.opt = opt self.hyp = hyp self.logger = logger # for printing results to console self.include = include self.keys = ['train/box_loss', 'train/obj_loss', 'train/cls_loss', # train loss 'metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', # metrics 'val/box_loss', 'val/obj_loss', 'val/cls_loss', # val loss 'x/lr0', 'x/lr1', 'x/lr2'] # params for k in LOGGERS: setattr(self, k, None) # init empty logger dictionary self.csv = True # always log to csv # Message if not wandb: prefix = colorstr('Weights & Biases: ') s = f"{prefix}run 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs (RECOMMENDED)" print(emojis(s)) # TensorBoard s = self.save_dir if 'tb' in self.include and not self.opt.evolve: prefix = colorstr('TensorBoard: ') self.logger.info(f"{prefix}Start with 'tensorboard --logdir {s.parent}', view at http://localhost:6006/") self.tb = SummaryWriter(str(s)) # W&B if wandb and 'wandb' in self.include: wandb_artifact_resume = isinstance(self.opt.resume, str) and self.opt.resume.startswith('wandb-artifact://') run_id = torch.load(self.weights).get('wandb_id') if self.opt.resume and not wandb_artifact_resume else None self.opt.hyp = self.hyp # add hyperparameters self.wandb = WandbLogger(self.opt, run_id) else: self.wandb = None def on_pretrain_routine_end(self): # Callback runs on pre-train routine end paths = self.save_dir.glob('*labels*.jpg') # training labels if self.wandb: self.wandb.log({"Labels": [wandb.Image(str(x), caption=x.name) for x in paths]}) def on_train_batch_end(self, ni, model, imgs, targets, paths, plots, sync_bn): # Callback runs on train batch end if plots: if ni == 0: if not sync_bn: # tb.add_graph() --sync known issue https://github.com/ultralytics/yolov5/issues/3754 with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress jit trace warning self.tb.add_graph(torch.jit.trace(de_parallel(model), imgs[0:1], strict=False), []) if ni < 3: f = self.save_dir / f'train_batch{ni}.jpg' # filename Thread(target=plot_images, args=(imgs, targets, paths, f), daemon=True).start() if self.wandb and ni == 10: files = sorted(self.save_dir.glob('train*.jpg')) self.wandb.log({'Mosaics': [wandb.Image(str(f), caption=f.name) for f in files if f.exists()]}) def on_train_epoch_end(self, epoch): # Callback runs on train epoch end if self.wandb: self.wandb.current_epoch = epoch + 1 def on_val_image_end(self, pred, predn, path, names, im): # Callback runs on val image end if self.wandb: self.wandb.val_one_image(pred, predn, path, names, im) def on_val_end(self): # Callback runs on val end if self.wandb: files = sorted(self.save_dir.glob('val*.jpg')) self.wandb.log({"Validation": [wandb.Image(str(f), caption=f.name) for f in files]}) def on_fit_epoch_end(self, vals, epoch, best_fitness, fi): # Callback runs at the end of each fit (train+val) epoch x = {k: v for k, v in zip(self.keys, vals)} # dict if self.csv: file = self.save_dir / 'results.csv' n = len(x) + 1 # number of cols s = '' if file.exists() else (('%20s,' * n % tuple(['epoch'] + self.keys)).rstrip(',') + '\n') # add header with open(file, 'a') as f: f.write(s + ('%20.5g,' * n % tuple([epoch] + vals)).rstrip(',') + '\n') if self.tb: for k, v in x.items(): self.tb.add_scalar(k, v, epoch) if self.wandb: self.wandb.log(x) self.wandb.end_epoch(best_result=best_fitness == fi) def on_model_save(self, last, epoch, final_epoch, best_fitness, fi): # Callback runs on model save event if self.wandb: if ((epoch + 1) % self.opt.save_period == 0 and not final_epoch) and self.opt.save_period != -1: self.wandb.log_model(last.parent, self.opt, epoch, fi, best_model=best_fitness == fi) def on_train_end(self, last, best, plots, epoch, results): # Callback runs on training end if plots: plot_results(file=self.save_dir / 'results.csv') # save results.png files = ['results.png', 'confusion_matrix.png', *(f'{x}_curve.png' for x in ('F1', 'PR', 'P', 'R'))] files = [(self.save_dir / f) for f in files if (self.save_dir / f).exists()] # filter if self.tb: import cv2 for f in files: self.tb.add_image(f.stem, cv2.imread(str(f))[..., ::-1], epoch, dataformats='HWC') if self.wandb: self.wandb.log({"Results": [wandb.Image(str(f), caption=f.name) for f in files]}) # Calling wandb.log. TODO: Refactor this into WandbLogger.log_model if not self.opt.evolve: wandb.log_artifact(str(best if best.exists() else last), type='model', name='run_' + self.wandb.wandb_run.id + '_model', aliases=['latest', 'best', 'stripped']) self.wandb.finish_run() else: self.wandb.finish_run() self.wandb = WandbLogger(self.opt) ================================================ FILE: utils/loggers/wandb/README.md ================================================ 📚 This guide explains how to use **Weights & Biases** (W&B) with YOLOv5 🚀. UPDATED 29 September 2021. * [About Weights & Biases](#about-weights-&-biases) * [First-Time Setup](#first-time-setup) * [Viewing runs](#viewing-runs) * [Disabling wandb](#disabling-wandb) * [Advanced Usage: Dataset Versioning and Evaluation](#advanced-usage) * [Reports: Share your work with the world!](#reports) ## About Weights & Biases Think of [W&B](https://wandb.ai/site?utm_campaign=repo_yolo_wandbtutorial) like GitHub for machine learning models. With a few lines of code, save everything you need to debug, compare and reproduce your models — architecture, hyperparameters, git commits, model weights, GPU usage, and even datasets and predictions. Used by top researchers including teams at OpenAI, Lyft, Github, and MILA, W&B is part of the new standard of best practices for machine learning. How W&B can help you optimize your machine learning workflows: * [Debug](https://wandb.ai/wandb/getting-started/reports/Visualize-Debug-Machine-Learning-Models--VmlldzoyNzY5MDk#Free-2) model performance in real time * [GPU usage](https://wandb.ai/wandb/getting-started/reports/Visualize-Debug-Machine-Learning-Models--VmlldzoyNzY5MDk#System-4) visualized automatically * [Custom charts](https://wandb.ai/wandb/customizable-charts/reports/Powerful-Custom-Charts-To-Debug-Model-Peformance--VmlldzoyNzY4ODI) for powerful, extensible visualization * [Share insights](https://wandb.ai/wandb/getting-started/reports/Visualize-Debug-Machine-Learning-Models--VmlldzoyNzY5MDk#Share-8) interactively with collaborators * [Optimize hyperparameters](https://docs.wandb.com/sweeps) efficiently * [Track](https://docs.wandb.com/artifacts) datasets, pipelines, and production models ## First-Time Setup
Toggle Details When you first train, W&B will prompt you to create a new account and will generate an **API key** for you. If you are an existing user you can retrieve your key from https://wandb.ai/authorize. This key is used to tell W&B where to log your data. You only need to supply your key once, and then it is remembered on the same device. W&B will create a cloud **project** (default is 'YOLOv5') for your training runs, and each new training run will be provided a unique run **name** within that project as project/name. You can also manually set your project and run name as: ```shell $ python train.py --project ... --name ... ``` YOLOv5 notebook example: Open In Colab Open In Kaggle Screen Shot 2021-09-29 at 10 23 13 PM
## Viewing Runs
Toggle Details Run information streams from your environment to the W&B cloud console as you train. This allows you to monitor and even cancel runs in realtime . All important information is logged: * Training & Validation losses * Metrics: Precision, Recall, mAP@0.5, mAP@0.5:0.95 * Learning Rate over time * A bounding box debugging panel, showing the training progress over time * GPU: Type, **GPU Utilization**, power, temperature, **CUDA memory usage** * System: Disk I/0, CPU utilization, RAM memory usage * Your trained model as W&B Artifact * Environment: OS and Python types, Git repository and state, **training command**

Weights & Biases dashboard

## Disabling wandb * training after running `wandb disabled` inside that directory creates no wandb run ![Screenshot (84)](https://user-images.githubusercontent.com/15766192/143441777-c780bdd7-7cb4-4404-9559-b4316030a985.png) * To enable wandb again, run `wandb online` ![Screenshot (85)](https://user-images.githubusercontent.com/15766192/143441866-7191b2cb-22f0-4e0f-ae64-2dc47dc13078.png) ## Advanced Usage You can leverage W&B artifacts and Tables integration to easily visualize and manage your datasets, models and training evaluations. Here are some quick examples to get you started.

1: Train and Log Evaluation simultaneousy

This is an extension of the previous section, but it'll also training after uploading the dataset. This also evaluation Table Evaluation table compares your predictions and ground truths across the validation set for each epoch. It uses the references to the already uploaded datasets, so no images will be uploaded from your system more than once.
Usage Code $ python train.py --upload_data val ![Screenshot from 2021-11-21 17-40-06](https://user-images.githubusercontent.com/15766192/142761183-c1696d8c-3f38-45ab-991a-bb0dfd98ae7d.png)

2. Visualize and Version Datasets

Log, visualize, dynamically query, and understand your data with W&B Tables. You can use the following command to log your dataset as a W&B Table. This will generate a {dataset}_wandb.yaml file which can be used to train from dataset artifact.
Usage Code $ python utils/logger/wandb/log_dataset.py --project ... --name ... --data .. ![Screenshot (64)](https://user-images.githubusercontent.com/15766192/128486078-d8433890-98a3-4d12-8986-b6c0e3fc64b9.png)

3: Train using dataset artifact

When you upload a dataset as described in the first section, you get a new config file with an added `_wandb` to its name. This file contains the information that can be used to train a model directly from the dataset artifact. This also logs evaluation
Usage Code $ python train.py --data {data}_wandb.yaml ![Screenshot (72)](https://user-images.githubusercontent.com/15766192/128979739-4cf63aeb-a76f-483f-8861-1c0100b938a5.png)

4: Save model checkpoints as artifacts

To enable saving and versioning checkpoints of your experiment, pass `--save_period n` with the base cammand, where `n` represents checkpoint interval. You can also log both the dataset and model checkpoints simultaneously. If not passed, only the final model will be logged
Usage Code $ python train.py --save_period 1 ![Screenshot (68)](https://user-images.githubusercontent.com/15766192/128726138-ec6c1f60-639d-437d-b4ee-3acd9de47ef3.png)

5: Resume runs from checkpoint artifacts.

Any run can be resumed using artifacts if the --resume argument starts with wandb-artifact:// prefix followed by the run path, i.e, wandb-artifact://username/project/runid . This doesn't require the model checkpoint to be present on the local system.
Usage Code $ python train.py --resume wandb-artifact://{run_path} ![Screenshot (70)](https://user-images.githubusercontent.com/15766192/128728988-4e84b355-6c87-41ae-a591-14aecf45343e.png)

6: Resume runs from dataset artifact & checkpoint artifacts.

Local dataset or model checkpoints are not required. This can be used to resume runs directly on a different device The syntax is same as the previous section, but you'll need to lof both the dataset and model checkpoints as artifacts, i.e, set bot --upload_dataset or train from _wandb.yaml file and set --save_period
Usage Code $ python train.py --resume wandb-artifact://{run_path} ![Screenshot (70)](https://user-images.githubusercontent.com/15766192/128728988-4e84b355-6c87-41ae-a591-14aecf45343e.png)

Reports

W&B Reports can be created from your saved runs for sharing online. Once a report is created you will receive a link you can use to publically share your results. Here is an example report created from the COCO128 tutorial trainings of all four YOLOv5 models ([link](https://wandb.ai/glenn-jocher/yolov5_tutorial/reports/YOLOv5-COCO128-Tutorial-Results--VmlldzozMDI5OTY)). Weights & Biases Reports ## Environments YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including [CUDA](https://developer.nvidia.com/cuda)/[CUDNN](https://developer.nvidia.com/cudnn), [Python](https://www.python.org/) and [PyTorch](https://pytorch.org/) preinstalled): - **Google Colab and Kaggle** notebooks with free GPU: Open In Colab Open In Kaggle - **Google Cloud** Deep Learning VM. See [GCP Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/GCP-Quickstart) - **Amazon** Deep Learning AMI. See [AWS Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/AWS-Quickstart) - **Docker Image**. See [Docker Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/Docker-Quickstart) Docker Pulls ## Status ![CI CPU testing](https://github.com/ultralytics/yolov5/workflows/CI%20CPU%20testing/badge.svg) If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), validation ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on MacOS, Windows, and Ubuntu every 24 hours and on every commit. ================================================ FILE: utils/loggers/wandb/__init__.py ================================================ ================================================ FILE: utils/loggers/wandb/log_dataset.py ================================================ import argparse from wandb_utils import WandbLogger from utils.general import LOGGER WANDB_ARTIFACT_PREFIX = 'wandb-artifact://' def create_dataset_artifact(opt): logger = WandbLogger(opt, None, job_type='Dataset Creation') # TODO: return value unused if not logger.wandb: LOGGER.info("install wandb using `pip install wandb` to log the dataset") if __name__ == '__main__': parser = argparse.ArgumentParser() parser.add_argument('--data', type=str, default='data/coco128.yaml', help='data.yaml path') parser.add_argument('--single-cls', action='store_true', help='train as single-class dataset') parser.add_argument('--project', type=str, default='YOLOv5', help='name of W&B Project') parser.add_argument('--entity', default=None, help='W&B entity') parser.add_argument('--name', type=str, default='log dataset', help='name of W&B run') opt = parser.parse_args() opt.resume = False # Explicitly disallow resume check for dataset upload job create_dataset_artifact(opt) ================================================ FILE: utils/loggers/wandb/sweep.py ================================================ import sys from pathlib import Path import wandb FILE = Path(__file__).resolve() ROOT = FILE.parents[3] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH from train import parse_opt, train from utils.callbacks import Callbacks from utils.general import increment_path from utils.torch_utils import select_device def sweep(): wandb.init() # Get hyp dict from sweep agent hyp_dict = vars(wandb.config).get("_items") # Workaround: get necessary opt args opt = parse_opt(known=True) opt.batch_size = hyp_dict.get("batch_size") opt.save_dir = str(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok or opt.evolve)) opt.epochs = hyp_dict.get("epochs") opt.nosave = True opt.data = hyp_dict.get("data") opt.weights = str(opt.weights) opt.cfg = str(opt.cfg) opt.data = str(opt.data) opt.hyp = str(opt.hyp) opt.project = str(opt.project) device = select_device(opt.device, batch_size=opt.batch_size) # train train(hyp_dict, opt, device, callbacks=Callbacks()) if __name__ == "__main__": sweep() ================================================ FILE: utils/loggers/wandb/sweep.yaml ================================================ # Hyperparameters for training # To set range- # Provide min and max values as: # parameter: # # min: scalar # max: scalar # OR # # Set a specific list of search space- # parameter: # values: [scalar1, scalar2, scalar3...] # # You can use grid, bayesian and hyperopt search strategy # For more info on configuring sweeps visit - https://docs.wandb.ai/guides/sweeps/configuration program: utils/loggers/wandb/sweep.py method: random metric: name: metrics/mAP_0.5 goal: maximize parameters: # hyperparameters: set either min, max range or values list data: value: "data/coco128.yaml" batch_size: values: [64] epochs: values: [10] lr0: distribution: uniform min: 1e-5 max: 1e-1 lrf: distribution: uniform min: 0.01 max: 1.0 momentum: distribution: uniform min: 0.6 max: 0.98 weight_decay: distribution: uniform min: 0.0 max: 0.001 warmup_epochs: distribution: uniform min: 0.0 max: 5.0 warmup_momentum: distribution: uniform min: 0.0 max: 0.95 warmup_bias_lr: distribution: uniform min: 0.0 max: 0.2 box: distribution: uniform min: 0.02 max: 0.2 cls: distribution: uniform min: 0.2 max: 4.0 cls_pw: distribution: uniform min: 0.5 max: 2.0 obj: distribution: uniform min: 0.2 max: 4.0 obj_pw: distribution: uniform min: 0.5 max: 2.0 iou_t: distribution: uniform min: 0.1 max: 0.7 anchor_t: distribution: uniform min: 2.0 max: 8.0 fl_gamma: distribution: uniform min: 0.0 max: 0.1 hsv_h: distribution: uniform min: 0.0 max: 0.1 hsv_s: distribution: uniform min: 0.0 max: 0.9 hsv_v: distribution: uniform min: 0.0 max: 0.9 degrees: distribution: uniform min: 0.0 max: 45.0 translate: distribution: uniform min: 0.0 max: 0.9 scale: distribution: uniform min: 0.0 max: 0.9 shear: distribution: uniform min: 0.0 max: 10.0 perspective: distribution: uniform min: 0.0 max: 0.001 flipud: distribution: uniform min: 0.0 max: 1.0 fliplr: distribution: uniform min: 0.0 max: 1.0 mosaic: distribution: uniform min: 0.0 max: 1.0 mixup: distribution: uniform min: 0.0 max: 1.0 copy_paste: distribution: uniform min: 0.0 max: 1.0 ================================================ FILE: utils/loggers/wandb/wandb_utils.py ================================================ """Utilities and tools for tracking runs with Weights & Biases.""" import logging import os import sys from contextlib import contextmanager from pathlib import Path from typing import Dict import yaml from tqdm import tqdm FILE = Path(__file__).resolve() ROOT = FILE.parents[3] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH from utils.datasets import LoadImagesAndLabels, img2label_paths from utils.general import LOGGER, check_dataset, check_file try: import wandb assert hasattr(wandb, '__version__') # verify package import not local dir except (ImportError, AssertionError): wandb = None RANK = int(os.getenv('RANK', -1)) WANDB_ARTIFACT_PREFIX = 'wandb-artifact://' def remove_prefix(from_string, prefix=WANDB_ARTIFACT_PREFIX): return from_string[len(prefix):] def check_wandb_config_file(data_config_file): wandb_config = '_wandb.'.join(data_config_file.rsplit('.', 1)) # updated data.yaml path if Path(wandb_config).is_file(): return wandb_config return data_config_file def check_wandb_dataset(data_file): is_trainset_wandb_artifact = False is_valset_wandb_artifact = False if check_file(data_file) and data_file.endswith('.yaml'): with open(data_file, errors='ignore') as f: data_dict = yaml.safe_load(f) is_trainset_wandb_artifact = (isinstance(data_dict['train'], str) and data_dict['train'].startswith(WANDB_ARTIFACT_PREFIX)) is_valset_wandb_artifact = (isinstance(data_dict['val'], str) and data_dict['val'].startswith(WANDB_ARTIFACT_PREFIX)) if is_trainset_wandb_artifact or is_valset_wandb_artifact: return data_dict else: return check_dataset(data_file) def get_run_info(run_path): run_path = Path(remove_prefix(run_path, WANDB_ARTIFACT_PREFIX)) run_id = run_path.stem project = run_path.parent.stem entity = run_path.parent.parent.stem model_artifact_name = 'run_' + run_id + '_model' return entity, project, run_id, model_artifact_name def check_wandb_resume(opt): process_wandb_config_ddp_mode(opt) if RANK not in [-1, 0] else None if isinstance(opt.resume, str): if opt.resume.startswith(WANDB_ARTIFACT_PREFIX): if RANK not in [-1, 0]: # For resuming DDP runs entity, project, run_id, model_artifact_name = get_run_info(opt.resume) api = wandb.Api() artifact = api.artifact(entity + '/' + project + '/' + model_artifact_name + ':latest') modeldir = artifact.download() opt.weights = str(Path(modeldir) / "last.pt") return True return None def process_wandb_config_ddp_mode(opt): with open(check_file(opt.data), errors='ignore') as f: data_dict = yaml.safe_load(f) # data dict train_dir, val_dir = None, None if isinstance(data_dict['train'], str) and data_dict['train'].startswith(WANDB_ARTIFACT_PREFIX): api = wandb.Api() train_artifact = api.artifact(remove_prefix(data_dict['train']) + ':' + opt.artifact_alias) train_dir = train_artifact.download() train_path = Path(train_dir) / 'data/images/' data_dict['train'] = str(train_path) if isinstance(data_dict['val'], str) and data_dict['val'].startswith(WANDB_ARTIFACT_PREFIX): api = wandb.Api() val_artifact = api.artifact(remove_prefix(data_dict['val']) + ':' + opt.artifact_alias) val_dir = val_artifact.download() val_path = Path(val_dir) / 'data/images/' data_dict['val'] = str(val_path) if train_dir or val_dir: ddp_data_path = str(Path(val_dir) / 'wandb_local_data.yaml') with open(ddp_data_path, 'w') as f: yaml.safe_dump(data_dict, f) opt.data = ddp_data_path class WandbLogger(): """Log training runs, datasets, models, and predictions to Weights & Biases. This logger sends information to W&B at wandb.ai. By default, this information includes hyperparameters, system configuration and metrics, model metrics, and basic data metrics and analyses. By providing additional command line arguments to train.py, datasets, models and predictions can also be logged. For more on how this logger is used, see the Weights & Biases documentation: https://docs.wandb.com/guides/integrations/yolov5 """ def __init__(self, opt, run_id=None, job_type='Training'): """ - Initialize WandbLogger instance - Upload dataset if opt.upload_dataset is True - Setup trainig processes if job_type is 'Training' arguments: opt (namespace) -- Commandline arguments for this run run_id (str) -- Run ID of W&B run to be resumed job_type (str) -- To set the job_type for this run """ # Pre-training routine -- self.job_type = job_type self.wandb, self.wandb_run = wandb, None if not wandb else wandb.run self.val_artifact, self.train_artifact = None, None self.train_artifact_path, self.val_artifact_path = None, None self.result_artifact = None self.val_table, self.result_table = None, None self.bbox_media_panel_images = [] self.val_table_path_map = None self.max_imgs_to_log = 16 self.wandb_artifact_data_dict = None self.data_dict = None # It's more elegant to stick to 1 wandb.init call, # but useful config data is overwritten in the WandbLogger's wandb.init call if isinstance(opt.resume, str): # checks resume from artifact if opt.resume.startswith(WANDB_ARTIFACT_PREFIX): entity, project, run_id, model_artifact_name = get_run_info(opt.resume) model_artifact_name = WANDB_ARTIFACT_PREFIX + model_artifact_name assert wandb, 'install wandb to resume wandb runs' # Resume wandb-artifact:// runs here| workaround for not overwriting wandb.config self.wandb_run = wandb.init(id=run_id, project=project, entity=entity, resume='allow', allow_val_change=True) opt.resume = model_artifact_name elif self.wandb: self.wandb_run = wandb.init(config=opt, resume="allow", project='YOLOv5' if opt.project == 'runs/train' else Path(opt.project).stem, entity=opt.entity, name=opt.name if opt.name != 'exp' else None, job_type=job_type, id=run_id, allow_val_change=True) if not wandb.run else wandb.run if self.wandb_run: if self.job_type == 'Training': if opt.upload_dataset: if not opt.resume: self.wandb_artifact_data_dict = self.check_and_upload_dataset(opt) if opt.resume: # resume from artifact if isinstance(opt.resume, str) and opt.resume.startswith(WANDB_ARTIFACT_PREFIX): self.data_dict = dict(self.wandb_run.config.data_dict) else: # local resume self.data_dict = check_wandb_dataset(opt.data) else: self.data_dict = check_wandb_dataset(opt.data) self.wandb_artifact_data_dict = self.wandb_artifact_data_dict or self.data_dict # write data_dict to config. useful for resuming from artifacts. Do this only when not resuming. self.wandb_run.config.update({'data_dict': self.wandb_artifact_data_dict}, allow_val_change=True) self.setup_training(opt) if self.job_type == 'Dataset Creation': self.wandb_run.config.update({"upload_dataset": True}) self.data_dict = self.check_and_upload_dataset(opt) def check_and_upload_dataset(self, opt): """ Check if the dataset format is compatible and upload it as W&B artifact arguments: opt (namespace)-- Commandline arguments for current run returns: Updated dataset info dictionary where local dataset paths are replaced by WAND_ARFACT_PREFIX links. """ assert wandb, 'Install wandb to upload dataset' config_path = self.log_dataset_artifact(opt.data, opt.single_cls, 'YOLOv5' if opt.project == 'runs/train' else Path(opt.project).stem) with open(config_path, errors='ignore') as f: wandb_data_dict = yaml.safe_load(f) return wandb_data_dict def setup_training(self, opt): """ Setup the necessary processes for training YOLO models: - Attempt to download model checkpoint and dataset artifacts if opt.resume stats with WANDB_ARTIFACT_PREFIX - Update data_dict, to contain info of previous run if resumed and the paths of dataset artifact if downloaded - Setup log_dict, initialize bbox_interval arguments: opt (namespace) -- commandline arguments for this run """ self.log_dict, self.current_epoch = {}, 0 self.bbox_interval = opt.bbox_interval if isinstance(opt.resume, str): modeldir, _ = self.download_model_artifact(opt) if modeldir: self.weights = Path(modeldir) / "last.pt" config = self.wandb_run.config opt.weights, opt.save_period, opt.batch_size, opt.bbox_interval, opt.epochs, opt.hyp = str( self.weights), config.save_period, config.batch_size, config.bbox_interval, config.epochs, \ config.hyp data_dict = self.data_dict if self.val_artifact is None: # If --upload_dataset is set, use the existing artifact, don't download self.train_artifact_path, self.train_artifact = self.download_dataset_artifact(data_dict.get('train'), opt.artifact_alias) self.val_artifact_path, self.val_artifact = self.download_dataset_artifact(data_dict.get('val'), opt.artifact_alias) if self.train_artifact_path is not None: train_path = Path(self.train_artifact_path) / 'data/images/' data_dict['train'] = str(train_path) if self.val_artifact_path is not None: val_path = Path(self.val_artifact_path) / 'data/images/' data_dict['val'] = str(val_path) if self.val_artifact is not None: self.result_artifact = wandb.Artifact("run_" + wandb.run.id + "_progress", "evaluation") columns = ["epoch", "id", "ground truth", "prediction"] columns.extend(self.data_dict['names']) self.result_table = wandb.Table(columns) self.val_table = self.val_artifact.get("val") if self.val_table_path_map is None: self.map_val_table_path() if opt.bbox_interval == -1: self.bbox_interval = opt.bbox_interval = (opt.epochs // 10) if opt.epochs > 10 else 1 train_from_artifact = self.train_artifact_path is not None and self.val_artifact_path is not None # Update the the data_dict to point to local artifacts dir if train_from_artifact: self.data_dict = data_dict def download_dataset_artifact(self, path, alias): """ download the model checkpoint artifact if the path starts with WANDB_ARTIFACT_PREFIX arguments: path -- path of the dataset to be used for training alias (str)-- alias of the artifact to be download/used for training returns: (str, wandb.Artifact) -- path of the downladed dataset and it's corresponding artifact object if dataset is found otherwise returns (None, None) """ if isinstance(path, str) and path.startswith(WANDB_ARTIFACT_PREFIX): artifact_path = Path(remove_prefix(path, WANDB_ARTIFACT_PREFIX) + ":" + alias) dataset_artifact = wandb.use_artifact(artifact_path.as_posix().replace("\\", "/")) assert dataset_artifact is not None, "'Error: W&B dataset artifact doesn\'t exist'" datadir = dataset_artifact.download() return datadir, dataset_artifact return None, None def download_model_artifact(self, opt): """ download the model checkpoint artifact if the resume path starts with WANDB_ARTIFACT_PREFIX arguments: opt (namespace) -- Commandline arguments for this run """ if opt.resume.startswith(WANDB_ARTIFACT_PREFIX): model_artifact = wandb.use_artifact(remove_prefix(opt.resume, WANDB_ARTIFACT_PREFIX) + ":latest") assert model_artifact is not None, 'Error: W&B model artifact doesn\'t exist' modeldir = model_artifact.download() epochs_trained = model_artifact.metadata.get('epochs_trained') total_epochs = model_artifact.metadata.get('total_epochs') is_finished = total_epochs is None assert not is_finished, 'training is finished, can only resume incomplete runs.' return modeldir, model_artifact return None, None def log_model(self, path, opt, epoch, fitness_score, best_model=False): """ Log the model checkpoint as W&B artifact arguments: path (Path) -- Path of directory containing the checkpoints opt (namespace) -- Command line arguments for this run epoch (int) -- Current epoch number fitness_score (float) -- fitness score for current epoch best_model (boolean) -- Boolean representing if the current checkpoint is the best yet. """ model_artifact = wandb.Artifact('run_' + wandb.run.id + '_model', type='model', metadata={ 'original_url': str(path), 'epochs_trained': epoch + 1, 'save period': opt.save_period, 'project': opt.project, 'total_epochs': opt.epochs, 'fitness_score': fitness_score }) model_artifact.add_file(str(path / 'last.pt'), name='last.pt') wandb.log_artifact(model_artifact, aliases=['latest', 'last', 'epoch ' + str(self.current_epoch), 'best' if best_model else '']) LOGGER.info(f"Saving model artifact on epoch {epoch + 1}") def log_dataset_artifact(self, data_file, single_cls, project, overwrite_config=False): """ Log the dataset as W&B artifact and return the new data file with W&B links arguments: data_file (str) -- the .yaml file with information about the dataset like - path, classes etc. single_class (boolean) -- train multi-class data as single-class project (str) -- project name. Used to construct the artifact path overwrite_config (boolean) -- overwrites the data.yaml file if set to true otherwise creates a new file with _wandb postfix. Eg -> data_wandb.yaml returns: the new .yaml file with artifact links. it can be used to start training directly from artifacts """ upload_dataset = self.wandb_run.config.upload_dataset log_val_only = isinstance(upload_dataset, str) and upload_dataset == 'val' self.data_dict = check_dataset(data_file) # parse and check data = dict(self.data_dict) nc, names = (1, ['item']) if single_cls else (int(data['nc']), data['names']) names = {k: v for k, v in enumerate(names)} # to index dictionary # log train set if not log_val_only: self.train_artifact = self.create_dataset_table(LoadImagesAndLabels( data['train'], rect=True, batch_size=1), names, name='train') if data.get('train') else None if data.get('train'): data['train'] = WANDB_ARTIFACT_PREFIX + str(Path(project) / 'train') self.val_artifact = self.create_dataset_table(LoadImagesAndLabels( data['val'], rect=True, batch_size=1), names, name='val') if data.get('val') else None if data.get('val'): data['val'] = WANDB_ARTIFACT_PREFIX + str(Path(project) / 'val') path = Path(data_file) # create a _wandb.yaml file with artifacts links if both train and test set are logged if not log_val_only: path = (path.stem if overwrite_config else path.stem + '_wandb') + '.yaml' # updated data.yaml path path = Path('data') / path data.pop('download', None) data.pop('path', None) with open(path, 'w') as f: yaml.safe_dump(data, f) LOGGER.info(f"Created dataset config file {path}") if self.job_type == 'Training': # builds correct artifact pipeline graph if not log_val_only: self.wandb_run.log_artifact( self.train_artifact) # calling use_artifact downloads the dataset. NOT NEEDED! self.wandb_run.use_artifact(self.val_artifact) self.val_artifact.wait() self.val_table = self.val_artifact.get('val') self.map_val_table_path() else: self.wandb_run.log_artifact(self.train_artifact) self.wandb_run.log_artifact(self.val_artifact) return path def map_val_table_path(self): """ Map the validation dataset Table like name of file -> it's id in the W&B Table. Useful for - referencing artifacts for evaluation. """ self.val_table_path_map = {} LOGGER.info("Mapping dataset") for i, data in enumerate(tqdm(self.val_table.data)): self.val_table_path_map[data[3]] = data[0] def create_dataset_table(self, dataset: LoadImagesAndLabels, class_to_id: Dict[int, str], name: str = 'dataset'): """ Create and return W&B artifact containing W&B Table of the dataset. arguments: dataset -- instance of LoadImagesAndLabels class used to iterate over the data to build Table class_to_id -- hash map that maps class ids to labels name -- name of the artifact returns: dataset artifact to be logged or used """ # TODO: Explore multiprocessing to slpit this loop parallely| This is essential for speeding up the the logging artifact = wandb.Artifact(name=name, type="dataset") img_files = tqdm([dataset.path]) if isinstance(dataset.path, str) and Path(dataset.path).is_dir() else None img_files = tqdm(dataset.img_files) if not img_files else img_files for img_file in img_files: if Path(img_file).is_dir(): artifact.add_dir(img_file, name='data/images') labels_path = 'labels'.join(dataset.path.rsplit('images', 1)) artifact.add_dir(labels_path, name='data/labels') else: artifact.add_file(img_file, name='data/images/' + Path(img_file).name) label_file = Path(img2label_paths([img_file])[0]) artifact.add_file(str(label_file), name='data/labels/' + label_file.name) if label_file.exists() else None table = wandb.Table(columns=["id", "train_image", "Classes", "name"]) class_set = wandb.Classes([{'id': id, 'name': name} for id, name in class_to_id.items()]) for si, (img, labels, paths, shapes) in enumerate(tqdm(dataset)): box_data, img_classes = [], {} for cls, *xywh in labels[:, 1:].tolist(): cls = int(cls) box_data.append({"position": {"middle": [xywh[0], xywh[1]], "width": xywh[2], "height": xywh[3]}, "class_id": cls, "box_caption": "%s" % (class_to_id[cls])}) img_classes[cls] = class_to_id[cls] boxes = {"ground_truth": {"box_data": box_data, "class_labels": class_to_id}} # inference-space table.add_data(si, wandb.Image(paths, classes=class_set, boxes=boxes), list(img_classes.values()), Path(paths).name) artifact.add(table, name) return artifact def log_training_progress(self, predn, path, names): """ Build evaluation Table. Uses reference from validation dataset table. arguments: predn (list): list of predictions in the native space in the format - [xmin, ymin, xmax, ymax, confidence, class] path (str): local path of the current evaluation image names (dict(int, str)): hash map that maps class ids to labels """ class_set = wandb.Classes([{'id': id, 'name': name} for id, name in names.items()]) box_data = [] avg_conf_per_class = [0] * len(self.data_dict['names']) pred_class_count = {} for *xyxy, conf, cls in predn.tolist(): if conf >= 0.25: cls = int(cls) box_data.append( {"position": {"minX": xyxy[0], "minY": xyxy[1], "maxX": xyxy[2], "maxY": xyxy[3]}, "class_id": cls, "box_caption": f"{names[cls]} {conf:.3f}", "scores": {"class_score": conf}, "domain": "pixel"}) avg_conf_per_class[cls] += conf if cls in pred_class_count: pred_class_count[cls] += 1 else: pred_class_count[cls] = 1 for pred_class in pred_class_count.keys(): avg_conf_per_class[pred_class] = avg_conf_per_class[pred_class] / pred_class_count[pred_class] boxes = {"predictions": {"box_data": box_data, "class_labels": names}} # inference-space id = self.val_table_path_map[Path(path).name] self.result_table.add_data(self.current_epoch, id, self.val_table.data[id][1], wandb.Image(self.val_table.data[id][1], boxes=boxes, classes=class_set), *avg_conf_per_class ) def val_one_image(self, pred, predn, path, names, im): """ Log validation data for one image. updates the result Table if validation dataset is uploaded and log bbox media panel arguments: pred (list): list of scaled predictions in the format - [xmin, ymin, xmax, ymax, confidence, class] predn (list): list of predictions in the native space - [xmin, ymin, xmax, ymax, confidence, class] path (str): local path of the current evaluation image """ if self.val_table and self.result_table: # Log Table if Val dataset is uploaded as artifact self.log_training_progress(predn, path, names) if len(self.bbox_media_panel_images) < self.max_imgs_to_log and self.current_epoch > 0: if self.current_epoch % self.bbox_interval == 0: box_data = [{"position": {"minX": xyxy[0], "minY": xyxy[1], "maxX": xyxy[2], "maxY": xyxy[3]}, "class_id": int(cls), "box_caption": f"{names[cls]} {conf:.3f}", "scores": {"class_score": conf}, "domain": "pixel"} for *xyxy, conf, cls in pred.tolist()] boxes = {"predictions": {"box_data": box_data, "class_labels": names}} # inference-space self.bbox_media_panel_images.append(wandb.Image(im, boxes=boxes, caption=path.name)) def log(self, log_dict): """ save the metrics to the logging dictionary arguments: log_dict (Dict) -- metrics/media to be logged in current step """ if self.wandb_run: for key, value in log_dict.items(): self.log_dict[key] = value def end_epoch(self, best_result=False): """ commit the log_dict, model artifacts and Tables to W&B and flush the log_dict. arguments: best_result (boolean): Boolean representing if the result of this evaluation is best or not """ if self.wandb_run: with all_logging_disabled(): if self.bbox_media_panel_images: self.log_dict["BoundingBoxDebugger"] = self.bbox_media_panel_images try: wandb.log(self.log_dict) except BaseException as e: LOGGER.info( f"An error occurred in wandb logger. The training will proceed without interruption. More info\n{e}") self.wandb_run.finish() self.wandb_run = None self.log_dict = {} self.bbox_media_panel_images = [] if self.result_artifact: self.result_artifact.add(self.result_table, 'result') wandb.log_artifact(self.result_artifact, aliases=['latest', 'last', 'epoch ' + str(self.current_epoch), ('best' if best_result else '')]) wandb.log({"evaluation": self.result_table}) columns = ["epoch", "id", "ground truth", "prediction"] columns.extend(self.data_dict['names']) self.result_table = wandb.Table(columns) self.result_artifact = wandb.Artifact("run_" + wandb.run.id + "_progress", "evaluation") def finish_run(self): """ Log metrics if any and finish the current W&B run """ if self.wandb_run: if self.log_dict: with all_logging_disabled(): wandb.log(self.log_dict) wandb.run.finish() @contextmanager def all_logging_disabled(highest_level=logging.CRITICAL): """ source - https://gist.github.com/simon-weber/7853144 A context manager that will prevent any logging messages triggered during the body from being processed. :param highest_level: the maximum logging level in use. This would only need to be changed if a custom level greater than CRITICAL is defined. """ previous_level = logging.root.manager.disable logging.disable(highest_level) try: yield finally: logging.disable(previous_level) ================================================ FILE: utils/loss.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Loss functions """ import torch import torch.nn as nn from utils.metrics import bbox_iou from utils.torch_utils import is_parallel def smooth_BCE(eps=0.1): # https://github.com/ultralytics/yolov3/issues/238#issuecomment-598028441 # return positive, negative label smoothing BCE targets return 1.0 - 0.5 * eps, 0.5 * eps class BCEBlurWithLogitsLoss(nn.Module): # BCEwithLogitLoss() with reduced missing label effects. def __init__(self, alpha=0.05): super().__init__() self.loss_fcn = nn.BCEWithLogitsLoss(reduction='none') # must be nn.BCEWithLogitsLoss() self.alpha = alpha def forward(self, pred, true): loss = self.loss_fcn(pred, true) pred = torch.sigmoid(pred) # prob from logits dx = pred - true # reduce only missing label effects # dx = (pred - true).abs() # reduce missing label and false label effects alpha_factor = 1 - torch.exp((dx - 1) / (self.alpha + 1e-4)) loss *= alpha_factor return loss.mean() class FocalLoss(nn.Module): # Wraps focal loss around existing loss_fcn(), i.e. criteria = FocalLoss(nn.BCEWithLogitsLoss(), gamma=1.5) def __init__(self, loss_fcn, gamma=1.5, alpha=0.25): super().__init__() self.loss_fcn = loss_fcn # must be nn.BCEWithLogitsLoss() self.gamma = gamma self.alpha = alpha self.reduction = loss_fcn.reduction self.loss_fcn.reduction = 'none' # required to apply FL to each element def forward(self, pred, true): loss = self.loss_fcn(pred, true) # p_t = torch.exp(-loss) # loss *= self.alpha * (1.000001 - p_t) ** self.gamma # non-zero power for gradient stability # TF implementation https://github.com/tensorflow/addons/blob/v0.7.1/tensorflow_addons/losses/focal_loss.py pred_prob = torch.sigmoid(pred) # prob from logits p_t = true * pred_prob + (1 - true) * (1 - pred_prob) alpha_factor = true * self.alpha + (1 - true) * (1 - self.alpha) modulating_factor = (1.0 - p_t) ** self.gamma loss *= alpha_factor * modulating_factor if self.reduction == 'mean': return loss.mean() elif self.reduction == 'sum': return loss.sum() else: # 'none' return loss class QFocalLoss(nn.Module): # Wraps Quality focal loss around existing loss_fcn(), i.e. criteria = FocalLoss(nn.BCEWithLogitsLoss(), gamma=1.5) def __init__(self, loss_fcn, gamma=1.5, alpha=0.25): super().__init__() self.loss_fcn = loss_fcn # must be nn.BCEWithLogitsLoss() self.gamma = gamma self.alpha = alpha self.reduction = loss_fcn.reduction self.loss_fcn.reduction = 'none' # required to apply FL to each element def forward(self, pred, true): loss = self.loss_fcn(pred, true) pred_prob = torch.sigmoid(pred) # prob from logits alpha_factor = true * self.alpha + (1 - true) * (1 - self.alpha) modulating_factor = torch.abs(true - pred_prob) ** self.gamma loss *= alpha_factor * modulating_factor if self.reduction == 'mean': return loss.mean() elif self.reduction == 'sum': return loss.sum() else: # 'none' return loss class ComputeLoss: # Compute losses def __init__(self, model, autobalance=False): self.sort_obj_iou = False device = next(model.parameters()).device # get model device h = model.hyp # hyperparameters # Define criteria BCEcls = nn.BCEWithLogitsLoss(pos_weight=torch.tensor([h['cls_pw']], device=device)) BCEobj = nn.BCEWithLogitsLoss(pos_weight=torch.tensor([h['obj_pw']], device=device)) # Class label smoothing https://arxiv.org/pdf/1902.04103.pdf eqn 3 self.cp, self.cn = smooth_BCE(eps=h.get('label_smoothing', 0.0)) # positive, negative BCE targets # Focal loss g = h['fl_gamma'] # focal loss gamma if g > 0: BCEcls, BCEobj = FocalLoss(BCEcls, g), FocalLoss(BCEobj, g) det = model.module.model[-1] if is_parallel(model) else model.model[-1] # Detect() module self.balance = {3: [4.0, 1.0, 0.4]}.get(det.nl, [4.0, 1.0, 0.25, 0.06, 0.02]) # P3-P7 self.ssi = list(det.stride).index(16) if autobalance else 0 # stride 16 index self.BCEcls, self.BCEobj, self.gr, self.hyp, self.autobalance = BCEcls, BCEobj, 1.0, h, autobalance for k in 'na', 'nc', 'nl', 'anchors': setattr(self, k, getattr(det, k)) def __call__(self, p, targets): # predictions, targets, model device = targets.device lcls, lbox, lobj = torch.zeros(1, device=device), torch.zeros(1, device=device), torch.zeros(1, device=device) tcls, tbox, indices, anchors = self.build_targets(p, targets) # targets # Losses for i, pi in enumerate(p): # layer index, layer predictions b, a, gj, gi = indices[i] # image, anchor, gridy, gridx tobj = torch.zeros_like(pi[..., 0], device=device) # target obj n = b.shape[0] # number of targets if n: ps = pi[b, a, gj, gi] # prediction subset corresponding to targets # Regression pxy = ps[:, :2].sigmoid() * 2 - 0.5 pwh = (ps[:, 2:4].sigmoid() * 2) ** 2 * anchors[i] pbox = torch.cat((pxy, pwh), 1) # predicted box iou = bbox_iou(pbox.T, tbox[i], x1y1x2y2=False, CIoU=True) # iou(prediction, target) lbox += (1.0 - iou).mean() # iou loss # Objectness score_iou = iou.detach().clamp(0).type(tobj.dtype) if self.sort_obj_iou: sort_id = torch.argsort(score_iou) b, a, gj, gi, score_iou = b[sort_id], a[sort_id], gj[sort_id], gi[sort_id], score_iou[sort_id] tobj[b, a, gj, gi] = (1.0 - self.gr) + self.gr * score_iou # iou ratio # Classification if self.nc > 1: # cls loss (only if multiple classes) t = torch.full_like(ps[:, 5:], self.cn, device=device) # targets t[range(n), tcls[i]] = self.cp lcls += self.BCEcls(ps[:, 5:], t) # BCE # Append targets to text file # with open('targets.txt', 'a') as file: # [file.write('%11.5g ' * 4 % tuple(x) + '\n') for x in torch.cat((txy[i], twh[i]), 1)] obji = self.BCEobj(pi[..., 4], tobj) lobj += obji * self.balance[i] # obj loss if self.autobalance: self.balance[i] = self.balance[i] * 0.9999 + 0.0001 / obji.detach().item() if self.autobalance: self.balance = [x / self.balance[self.ssi] for x in self.balance] lbox *= self.hyp['box'] lobj *= self.hyp['obj'] lcls *= self.hyp['cls'] bs = tobj.shape[0] # batch size return (lbox + lobj + lcls) * bs, torch.cat((lbox, lobj, lcls)).detach() def build_targets(self, p, targets): # Build targets for compute_loss(), input targets(image,class,x,y,w,h) na, nt = self.na, targets.shape[0] # number of anchors, targets tcls, tbox, indices, anch = [], [], [], [] gain = torch.ones(7, device=targets.device) # normalized to gridspace gain ai = torch.arange(na, device=targets.device).float().view(na, 1).repeat(1, nt) # same as .repeat_interleave(nt) targets = torch.cat((targets.repeat(na, 1, 1), ai[:, :, None]), 2) # append anchor indices g = 0.5 # bias off = torch.tensor([[0, 0], [1, 0], [0, 1], [-1, 0], [0, -1], # j,k,l,m # [1, 1], [1, -1], [-1, 1], [-1, -1], # jk,jm,lk,lm ], device=targets.device).float() * g # offsets for i in range(self.nl): anchors = self.anchors[i] gain[2:6] = torch.tensor(p[i].shape)[[3, 2, 3, 2]] # xyxy gain # Match targets to anchors t = targets * gain if nt: # Matches r = t[:, :, 4:6] / anchors[:, None] # wh ratio j = torch.max(r, 1 / r).max(2)[0] < self.hyp['anchor_t'] # compare # j = wh_iou(anchors, t[:, 4:6]) > model.hyp['iou_t'] # iou(3,n)=wh_iou(anchors(3,2), gwh(n,2)) t = t[j] # filter # Offsets gxy = t[:, 2:4] # grid xy gxi = gain[[2, 3]] - gxy # inverse j, k = ((gxy % 1 < g) & (gxy > 1)).T l, m = ((gxi % 1 < g) & (gxi > 1)).T j = torch.stack((torch.ones_like(j), j, k, l, m)) t = t.repeat((5, 1, 1))[j] offsets = (torch.zeros_like(gxy)[None] + off[:, None])[j] else: t = targets[0] offsets = 0 # Define b, c = t[:, :2].long().T # image, class gxy = t[:, 2:4] # grid xy gwh = t[:, 4:6] # grid wh gij = (gxy - offsets).long() gi, gj = gij.T # grid xy indices # Append a = t[:, 6].long() # anchor indices indices.append((b, a, gj.clamp_(0, gain[3] - 1), gi.clamp_(0, gain[2] - 1))) # image, anchor, grid indices tbox.append(torch.cat((gxy - gij, gwh), 1)) # box anch.append(anchors[a]) # anchors tcls.append(c) # class return tcls, tbox, indices, anch ================================================ FILE: utils/metrics.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Model validation metrics """ import math import warnings from pathlib import Path import matplotlib.pyplot as plt import numpy as np import torch def fitness(x): # Model fitness as a weighted combination of metrics w = [0.0, 0.0, 0.1, 0.9] # weights for [P, R, mAP@0.5, mAP@0.5:0.95] return (x[:, :4] * w).sum(1) def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='.', names=(), eps=1e-16): """ Compute the average precision, given the recall and precision curves. Source: https://github.com/rafaelpadilla/Object-Detection-Metrics. # Arguments tp: True positives (nparray, nx1 or nx10). conf: Objectness value from 0-1 (nparray). pred_cls: Predicted object classes (nparray). target_cls: True object classes (nparray). plot: Plot precision-recall curve at mAP@0.5 save_dir: Plot save directory # Returns The average precision as computed in py-faster-rcnn. """ # Sort by objectness i = np.argsort(-conf) tp, conf, pred_cls = tp[i], conf[i], pred_cls[i] # Find unique classes unique_classes, nt = np.unique(target_cls, return_counts=True) nc = unique_classes.shape[0] # number of classes, number of detections # Create Precision-Recall curve and compute AP for each class px, py = np.linspace(0, 1, 1000), [] # for plotting ap, p, r = np.zeros((nc, tp.shape[1])), np.zeros((nc, 1000)), np.zeros((nc, 1000)) for ci, c in enumerate(unique_classes): i = pred_cls == c n_l = nt[ci] # number of labels n_p = i.sum() # number of predictions if n_p == 0 or n_l == 0: continue else: # Accumulate FPs and TPs fpc = (1 - tp[i]).cumsum(0) tpc = tp[i].cumsum(0) # Recall recall = tpc / (n_l + eps) # recall curve r[ci] = np.interp(-px, -conf[i], recall[:, 0], left=0) # negative x, xp because xp decreases # Precision precision = tpc / (tpc + fpc) # precision curve p[ci] = np.interp(-px, -conf[i], precision[:, 0], left=1) # p at pr_score # AP from recall-precision curve for j in range(tp.shape[1]): ap[ci, j], mpre, mrec = compute_ap(recall[:, j], precision[:, j]) if plot and j == 0: py.append(np.interp(px, mrec, mpre)) # precision at mAP@0.5 # Compute F1 (harmonic mean of precision and recall) f1 = 2 * p * r / (p + r + eps) names = [v for k, v in names.items() if k in unique_classes] # list: only classes that have data names = {i: v for i, v in enumerate(names)} # to dict if plot: plot_pr_curve(px, py, ap, Path(save_dir) / 'PR_curve.png', names) plot_mc_curve(px, f1, Path(save_dir) / 'F1_curve.png', names, ylabel='F1') plot_mc_curve(px, p, Path(save_dir) / 'P_curve.png', names, ylabel='Precision') plot_mc_curve(px, r, Path(save_dir) / 'R_curve.png', names, ylabel='Recall') i = f1.mean(0).argmax() # max F1 index p, r, f1 = p[:, i], r[:, i], f1[:, i] tp = (r * nt).round() # true positives fp = (tp / (p + eps) - tp).round() # false positives return tp, fp, p, r, f1, ap, unique_classes.astype('int32') def compute_ap(recall, precision): """ Compute the average precision, given the recall and precision curves # Arguments recall: The recall curve (list) precision: The precision curve (list) # Returns Average precision, precision curve, recall curve """ # Append sentinel values to beginning and end mrec = np.concatenate(([0.0], recall, [1.0])) mpre = np.concatenate(([1.0], precision, [0.0])) # Compute the precision envelope mpre = np.flip(np.maximum.accumulate(np.flip(mpre))) # Integrate area under curve method = 'interp' # methods: 'continuous', 'interp' if method == 'interp': x = np.linspace(0, 1, 101) # 101-point interp (COCO) ap = np.trapz(np.interp(x, mrec, mpre), x) # integrate else: # 'continuous' i = np.where(mrec[1:] != mrec[:-1])[0] # points where x axis (recall) changes ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1]) # area under curve return ap, mpre, mrec class ConfusionMatrix: # Updated version of https://github.com/kaanakan/object_detection_confusion_matrix def __init__(self, nc, conf=0.25, iou_thres=0.45): self.matrix = np.zeros((nc + 1, nc + 1)) self.nc = nc # number of classes self.conf = conf self.iou_thres = iou_thres def process_batch(self, detections, labels): """ Return intersection-over-union (Jaccard index) of boxes. Both sets of boxes are expected to be in (x1, y1, x2, y2) format. Arguments: detections (Array[N, 6]), x1, y1, x2, y2, conf, class labels (Array[M, 5]), class, x1, y1, x2, y2 Returns: None, updates confusion matrix accordingly """ detections = detections[detections[:, 4] > self.conf] gt_classes = labels[:, 0].int() detection_classes = detections[:, 5].int() iou = box_iou(labels[:, 1:], detections[:, :4]) x = torch.where(iou > self.iou_thres) if x[0].shape[0]: matches = torch.cat((torch.stack(x, 1), iou[x[0], x[1]][:, None]), 1).cpu().numpy() if x[0].shape[0] > 1: matches = matches[matches[:, 2].argsort()[::-1]] matches = matches[np.unique(matches[:, 1], return_index=True)[1]] matches = matches[matches[:, 2].argsort()[::-1]] matches = matches[np.unique(matches[:, 0], return_index=True)[1]] else: matches = np.zeros((0, 3)) n = matches.shape[0] > 0 m0, m1, _ = matches.transpose().astype(np.int16) for i, gc in enumerate(gt_classes): j = m0 == i if n and sum(j) == 1: self.matrix[detection_classes[m1[j]], gc] += 1 # correct else: self.matrix[self.nc, gc] += 1 # background FP if n: for i, dc in enumerate(detection_classes): if not any(m1 == i): self.matrix[dc, self.nc] += 1 # background FN def matrix(self): return self.matrix def tp_fp(self): tp = self.matrix.diagonal() # true positives fp = self.matrix.sum(1) - tp # false positives # fn = self.matrix.sum(0) - tp # false negatives (missed detections) return tp[:-1], fp[:-1] # remove background class def plot(self, normalize=True, save_dir='', names=()): try: import seaborn as sn array = self.matrix / ((self.matrix.sum(0).reshape(1, -1) + 1E-6) if normalize else 1) # normalize columns array[array < 0.005] = np.nan # don't annotate (would appear as 0.00) fig = plt.figure(figsize=(12, 9), tight_layout=True) sn.set(font_scale=1.0 if self.nc < 50 else 0.8) # for label size labels = (0 < len(names) < 99) and len(names) == self.nc # apply names to ticklabels with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress empty matrix RuntimeWarning: All-NaN slice encountered sn.heatmap(array, annot=self.nc < 30, annot_kws={"size": 8}, cmap='Blues', fmt='.2f', square=True, xticklabels=names + ['background FP'] if labels else "auto", yticklabels=names + ['background FN'] if labels else "auto").set_facecolor((1, 1, 1)) fig.axes[0].set_xlabel('True') fig.axes[0].set_ylabel('Predicted') fig.savefig(Path(save_dir) / 'confusion_matrix.png', dpi=250) plt.close() except Exception as e: print(f'WARNING: ConfusionMatrix plot failure: {e}') def print(self): for i in range(self.nc + 1): print(' '.join(map(str, self.matrix[i]))) def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7): # Returns the IoU of box1 to box2. box1 is 4, box2 is nx4 box2 = box2.T # Get the coordinates of bounding boxes if x1y1x2y2: # x1, y1, x2, y2 = box1 b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3] b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3] else: # transform from xywh to xyxy b1_x1, b1_x2 = box1[0] - box1[2] / 2, box1[0] + box1[2] / 2 b1_y1, b1_y2 = box1[1] - box1[3] / 2, box1[1] + box1[3] / 2 b2_x1, b2_x2 = box2[0] - box2[2] / 2, box2[0] + box2[2] / 2 b2_y1, b2_y2 = box2[1] - box2[3] / 2, box2[1] + box2[3] / 2 # Intersection area inter = (torch.min(b1_x2, b2_x2) - torch.max(b1_x1, b2_x1)).clamp(0) * \ (torch.min(b1_y2, b2_y2) - torch.max(b1_y1, b2_y1)).clamp(0) # Union Area w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps union = w1 * h1 + w2 * h2 - inter + eps iou = inter / union if GIoU or DIoU or CIoU: cw = torch.max(b1_x2, b2_x2) - torch.min(b1_x1, b2_x1) # convex (smallest enclosing box) width ch = torch.max(b1_y2, b2_y2) - torch.min(b1_y1, b2_y1) # convex height if CIoU or DIoU: # Distance or Complete IoU https://arxiv.org/abs/1911.08287v1 c2 = cw ** 2 + ch ** 2 + eps # convex diagonal squared rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 + (b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4 # center distance squared if DIoU: return iou - rho2 / c2 # DIoU elif CIoU: # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47 v = (4 / math.pi ** 2) * torch.pow(torch.atan(w2 / h2) - torch.atan(w1 / h1), 2) with torch.no_grad(): alpha = v / (v - iou + (1 + eps)) return iou - (rho2 / c2 + v * alpha) # CIoU else: # GIoU https://arxiv.org/pdf/1902.09630.pdf c_area = cw * ch + eps # convex area return iou - (c_area - union) / c_area # GIoU else: return iou # IoU def box_iou(box1, box2): # https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py """ Return intersection-over-union (Jaccard index) of boxes. Both sets of boxes are expected to be in (x1, y1, x2, y2) format. Arguments: box1 (Tensor[N, 4]) box2 (Tensor[M, 4]) Returns: iou (Tensor[N, M]): the NxM matrix containing the pairwise IoU values for every element in boxes1 and boxes2 """ def box_area(box): # box = 4xn return (box[2] - box[0]) * (box[3] - box[1]) area1 = box_area(box1.T) area2 = box_area(box2.T) # inter(N,M) = (rb(N,M,2) - lt(N,M,2)).clamp(0).prod(2) inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2) return inter / (area1[:, None] + area2 - inter) # iou = inter / (area1 + area2 - inter) def bbox_ioa(box1, box2, eps=1E-7): """ Returns the intersection over box2 area given box1, box2. Boxes are x1y1x2y2 box1: np.array of shape(4) box2: np.array of shape(nx4) returns: np.array of shape(n) """ box2 = box2.transpose() # Get the coordinates of bounding boxes b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3] b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3] # Intersection area inter_area = (np.minimum(b1_x2, b2_x2) - np.maximum(b1_x1, b2_x1)).clip(0) * \ (np.minimum(b1_y2, b2_y2) - np.maximum(b1_y1, b2_y1)).clip(0) # box2 area box2_area = (b2_x2 - b2_x1) * (b2_y2 - b2_y1) + eps # Intersection over box2 area return inter_area / box2_area def wh_iou(wh1, wh2): # Returns the nxm IoU matrix. wh1 is nx2, wh2 is mx2 wh1 = wh1[:, None] # [N,1,2] wh2 = wh2[None] # [1,M,2] inter = torch.min(wh1, wh2).prod(2) # [N,M] return inter / (wh1.prod(2) + wh2.prod(2) - inter) # iou = inter / (area1 + area2 - inter) # Plots ---------------------------------------------------------------------------------------------------------------- def plot_pr_curve(px, py, ap, save_dir='pr_curve.png', names=()): # Precision-recall curve fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True) py = np.stack(py, axis=1) if 0 < len(names) < 21: # display per-class legend if < 21 classes for i, y in enumerate(py.T): ax.plot(px, y, linewidth=1, label=f'{names[i]} {ap[i, 0]:.3f}') # plot(recall, precision) else: ax.plot(px, py, linewidth=1, color='grey') # plot(recall, precision) ax.plot(px, py.mean(1), linewidth=3, color='blue', label='all classes %.3f mAP@0.5' % ap[:, 0].mean()) ax.set_xlabel('Recall') ax.set_ylabel('Precision') ax.set_xlim(0, 1) ax.set_ylim(0, 1) plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left") fig.savefig(Path(save_dir), dpi=250) plt.close() def plot_mc_curve(px, py, save_dir='mc_curve.png', names=(), xlabel='Confidence', ylabel='Metric'): # Metric-confidence curve fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True) if 0 < len(names) < 21: # display per-class legend if < 21 classes for i, y in enumerate(py): ax.plot(px, y, linewidth=1, label=f'{names[i]}') # plot(confidence, metric) else: ax.plot(px, py.T, linewidth=1, color='grey') # plot(confidence, metric) y = py.mean(0) ax.plot(px, y, linewidth=3, color='blue', label=f'all classes {y.max():.2f} at {px[y.argmax()]:.3f}') ax.set_xlabel(xlabel) ax.set_ylabel(ylabel) ax.set_xlim(0, 1) ax.set_ylim(0, 1) plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left") fig.savefig(Path(save_dir), dpi=250) plt.close() ================================================ FILE: utils/plots.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Plotting utils """ import math import os from copy import copy from pathlib import Path import cv2 import matplotlib import matplotlib.pyplot as plt import numpy as np import pandas as pd import seaborn as sn import torch from PIL import Image, ImageDraw, ImageFont from utils.general import (LOGGER, Timeout, check_requirements, clip_coords, increment_path, is_ascii, is_chinese, try_except, user_config_dir, xywh2xyxy, xyxy2xywh) from utils.metrics import fitness # Settings CONFIG_DIR = user_config_dir() # Ultralytics settings dir RANK = int(os.getenv('RANK', -1)) matplotlib.rc('font', **{'size': 11}) matplotlib.use('Agg') # for writing to files only class Colors: # Ultralytics color palette https://ultralytics.com/ def __init__(self): # hex = matplotlib.colors.TABLEAU_COLORS.values() hex = ('FF3838', 'FF9D97', 'FF701F', 'FFB21D', 'CFD231', '48F90A', '92CC17', '3DDB86', '1A9334', '00D4BB', '2C99A8', '00C2FF', '344593', '6473FF', '0018EC', '8438FF', '520085', 'CB38FF', 'FF95C8', 'FF37C7') self.palette = [self.hex2rgb('#' + c) for c in hex] self.n = len(self.palette) def __call__(self, i, bgr=False): c = self.palette[int(i) % self.n] return (c[2], c[1], c[0]) if bgr else c @staticmethod def hex2rgb(h): # rgb order (PIL) return tuple(int(h[1 + i:1 + i + 2], 16) for i in (0, 2, 4)) colors = Colors() # create instance for 'from utils.plots import colors' def check_font(font='Arial.ttf', size=10): # Return a PIL TrueType Font, downloading to CONFIG_DIR if necessary font = Path(font) font = font if font.exists() else (CONFIG_DIR / font.name) try: return ImageFont.truetype(str(font) if font.exists() else font.name, size) except Exception as e: # download if missing url = "https://ultralytics.com/assets/" + font.name print(f'Downloading {url} to {font}...') torch.hub.download_url_to_file(url, str(font), progress=False) try: return ImageFont.truetype(str(font), size) except TypeError: check_requirements('Pillow>=8.4.0') # known issue https://github.com/ultralytics/yolov5/issues/5374 class Annotator: if RANK in (-1, 0): check_font() # download TTF if necessary # YOLOv5 Annotator for train/val mosaics and jpgs and detect/hub inference annotations def __init__(self, im, line_width=None, font_size=None, font='Arial.ttf', pil=False, example='abc'): assert im.data.contiguous, 'Image not contiguous. Apply np.ascontiguousarray(im) to Annotator() input images.' self.pil = pil or not is_ascii(example) or is_chinese(example) if self.pil: # use PIL self.im = im if isinstance(im, Image.Image) else Image.fromarray(im) self.draw = ImageDraw.Draw(self.im) self.font = check_font(font='Arial.Unicode.ttf' if is_chinese(example) else font, size=font_size or max(round(sum(self.im.size) / 2 * 0.035), 12)) else: # use cv2 self.im = im self.lw = line_width or max(round(sum(im.shape) / 2 * 0.003), 2) # line width def box_label(self, box, label='', color=(128, 128, 128), txt_color=(255, 255, 255)): # Add one xyxy box to image with label if self.pil or not is_ascii(label): self.draw.rectangle(box, width=self.lw, outline=color) # box if label: w, h = self.font.getsize(label) # text width, height outside = box[1] - h >= 0 # label fits outside box self.draw.rectangle([box[0], box[1] - h if outside else box[1], box[0] + w + 1, box[1] + 1 if outside else box[1] + h + 1], fill=color) # self.draw.text((box[0], box[1]), label, fill=txt_color, font=self.font, anchor='ls') # for PIL>8.0 self.draw.text((box[0], box[1] - h if outside else box[1]), label, fill=txt_color, font=self.font) else: # cv2 p1, p2 = (int(box[0]), int(box[1])), (int(box[2]), int(box[3])) cv2.rectangle(self.im, p1, p2, color, thickness=self.lw, lineType=cv2.LINE_AA) if label: tf = max(self.lw - 1, 1) # font thickness w, h = cv2.getTextSize(label, 0, fontScale=self.lw / 3, thickness=tf)[0] # text width, height outside = p1[1] - h - 3 >= 0 # label fits outside box p2 = p1[0] + w, p1[1] - h - 3 if outside else p1[1] + h + 3 cv2.rectangle(self.im, p1, p2, color, -1, cv2.LINE_AA) # filled cv2.putText(self.im, label, (p1[0], p1[1] - 2 if outside else p1[1] + h + 2), 0, self.lw / 3, txt_color, thickness=tf, lineType=cv2.LINE_AA) def rectangle(self, xy, fill=None, outline=None, width=1): # Add rectangle to image (PIL-only) self.draw.rectangle(xy, fill, outline, width) def text(self, xy, text, txt_color=(255, 255, 255)): # Add text to image (PIL-only) w, h = self.font.getsize(text) # text width, height self.draw.text((xy[0], xy[1] - h + 1), text, fill=txt_color, font=self.font) def result(self): # Return annotated image as array return np.asarray(self.im) def feature_visualization(x, module_type, stage, n=32, save_dir=Path('runs/detect/exp')): """ x: Features to be visualized module_type: Module type stage: Module stage within model n: Maximum number of feature maps to plot save_dir: Directory to save results """ if 'Detect' not in module_type: batch, channels, height, width = x.shape # batch, channels, height, width if height > 1 and width > 1: f = save_dir / f"stage{stage}_{module_type.split('.')[-1]}_features.png" # filename blocks = torch.chunk(x[0].cpu(), channels, dim=0) # select batch index 0, block by channels n = min(n, channels) # number of plots fig, ax = plt.subplots(math.ceil(n / 8), 8, tight_layout=True) # 8 rows x n/8 cols ax = ax.ravel() plt.subplots_adjust(wspace=0.05, hspace=0.05) for i in range(n): ax[i].imshow(blocks[i].squeeze()) # cmap='gray' ax[i].axis('off') print(f'Saving {f}... ({n}/{channels})') plt.savefig(f, dpi=300, bbox_inches='tight') plt.close() np.save(str(f.with_suffix('.npy')), x[0].cpu().numpy()) # npy save def hist2d(x, y, n=100): # 2d histogram used in labels.png and evolve.png xedges, yedges = np.linspace(x.min(), x.max(), n), np.linspace(y.min(), y.max(), n) hist, xedges, yedges = np.histogram2d(x, y, (xedges, yedges)) xidx = np.clip(np.digitize(x, xedges) - 1, 0, hist.shape[0] - 1) yidx = np.clip(np.digitize(y, yedges) - 1, 0, hist.shape[1] - 1) return np.log(hist[xidx, yidx]) def butter_lowpass_filtfilt(data, cutoff=1500, fs=50000, order=5): from scipy.signal import butter, filtfilt # https://stackoverflow.com/questions/28536191/how-to-filter-smooth-with-scipy-numpy def butter_lowpass(cutoff, fs, order): nyq = 0.5 * fs normal_cutoff = cutoff / nyq return butter(order, normal_cutoff, btype='low', analog=False) b, a = butter_lowpass(cutoff, fs, order=order) return filtfilt(b, a, data) # forward-backward filter def output_to_target(output): # Convert model output to target format [batch_id, class_id, x, y, w, h, conf] targets = [] for i, o in enumerate(output): for *box, conf, cls in o.cpu().numpy(): targets.append([i, cls, *list(*xyxy2xywh(np.array(box)[None])), conf]) return np.array(targets) def plot_images(images, targets, paths=None, fname='images.jpg', names=None, max_size=1920, max_subplots=16): # Plot image grid with labels if isinstance(images, torch.Tensor): images = images.cpu().float().numpy() if isinstance(targets, torch.Tensor): targets = targets.cpu().numpy() if np.max(images[0]) <= 1: images *= 255 # de-normalise (optional) bs, _, h, w = images.shape # batch size, _, height, width bs = min(bs, max_subplots) # limit plot images ns = np.ceil(bs ** 0.5) # number of subplots (square) # Build Image mosaic = np.full((int(ns * h), int(ns * w), 3), 255, dtype=np.uint8) # init for i, im in enumerate(images): if i == max_subplots: # if last batch has fewer images than we expect break x, y = int(w * (i // ns)), int(h * (i % ns)) # block origin im = im.transpose(1, 2, 0) mosaic[y:y + h, x:x + w, :] = im # Resize (optional) scale = max_size / ns / max(h, w) if scale < 1: h = math.ceil(scale * h) w = math.ceil(scale * w) mosaic = cv2.resize(mosaic, tuple(int(x * ns) for x in (w, h))) # Annotate fs = int((h + w) * ns * 0.01) # font size annotator = Annotator(mosaic, line_width=round(fs / 10), font_size=fs, pil=True) for i in range(i + 1): x, y = int(w * (i // ns)), int(h * (i % ns)) # block origin annotator.rectangle([x, y, x + w, y + h], None, (255, 255, 255), width=2) # borders if paths: annotator.text((x + 5, y + 5 + h), text=Path(paths[i]).name[:40], txt_color=(220, 220, 220)) # filenames if len(targets) > 0: ti = targets[targets[:, 0] == i] # image targets boxes = xywh2xyxy(ti[:, 2:6]).T classes = ti[:, 1].astype('int') labels = ti.shape[1] == 6 # labels if no conf column conf = None if labels else ti[:, 6] # check for confidence presence (label vs pred) if boxes.shape[1]: if boxes.max() <= 1.01: # if normalized with tolerance 0.01 boxes[[0, 2]] *= w # scale to pixels boxes[[1, 3]] *= h elif scale < 1: # absolute coords need scale if image scales boxes *= scale boxes[[0, 2]] += x boxes[[1, 3]] += y for j, box in enumerate(boxes.T.tolist()): cls = classes[j] color = colors(cls) cls = names[cls] if names else cls if labels or conf[j] > 0.25: # 0.25 conf thresh label = f'{cls}' if labels else f'{cls} {conf[j]:.1f}' annotator.box_label(box, label, color=color) annotator.im.save(fname) # save def plot_lr_scheduler(optimizer, scheduler, epochs=300, save_dir=''): # Plot LR simulating training for full epochs optimizer, scheduler = copy(optimizer), copy(scheduler) # do not modify originals y = [] for _ in range(epochs): scheduler.step() y.append(optimizer.param_groups[0]['lr']) plt.plot(y, '.-', label='LR') plt.xlabel('epoch') plt.ylabel('LR') plt.grid() plt.xlim(0, epochs) plt.ylim(0) plt.savefig(Path(save_dir) / 'LR.png', dpi=200) plt.close() def plot_val_txt(): # from utils.plots import *; plot_val() # Plot val.txt histograms x = np.loadtxt('val.txt', dtype=np.float32) box = xyxy2xywh(x[:, :4]) cx, cy = box[:, 0], box[:, 1] fig, ax = plt.subplots(1, 1, figsize=(6, 6), tight_layout=True) ax.hist2d(cx, cy, bins=600, cmax=10, cmin=0) ax.set_aspect('equal') plt.savefig('hist2d.png', dpi=300) fig, ax = plt.subplots(1, 2, figsize=(12, 6), tight_layout=True) ax[0].hist(cx, bins=600) ax[1].hist(cy, bins=600) plt.savefig('hist1d.png', dpi=200) def plot_targets_txt(): # from utils.plots import *; plot_targets_txt() # Plot targets.txt histograms x = np.loadtxt('targets.txt', dtype=np.float32).T s = ['x targets', 'y targets', 'width targets', 'height targets'] fig, ax = plt.subplots(2, 2, figsize=(8, 8), tight_layout=True) ax = ax.ravel() for i in range(4): ax[i].hist(x[i], bins=100, label=f'{x[i].mean():.3g} +/- {x[i].std():.3g}') ax[i].legend() ax[i].set_title(s[i]) plt.savefig('targets.jpg', dpi=200) def plot_val_study(file='', dir='', x=None): # from utils.plots import *; plot_val_study() # Plot file=study.txt generated by val.py (or plot all study*.txt in dir) save_dir = Path(file).parent if file else Path(dir) plot2 = False # plot additional results if plot2: ax = plt.subplots(2, 4, figsize=(10, 6), tight_layout=True)[1].ravel() fig2, ax2 = plt.subplots(1, 1, figsize=(8, 4), tight_layout=True) # for f in [save_dir / f'study_coco_{x}.txt' for x in ['yolov5n6', 'yolov5s6', 'yolov5m6', 'yolov5l6', 'yolov5x6']]: for f in sorted(save_dir.glob('study*.txt')): y = np.loadtxt(f, dtype=np.float32, usecols=[0, 1, 2, 3, 7, 8, 9], ndmin=2).T x = np.arange(y.shape[1]) if x is None else np.array(x) if plot2: s = ['P', 'R', 'mAP@.5', 'mAP@.5:.95', 't_preprocess (ms/img)', 't_inference (ms/img)', 't_NMS (ms/img)'] for i in range(7): ax[i].plot(x, y[i], '.-', linewidth=2, markersize=8) ax[i].set_title(s[i]) j = y[3].argmax() + 1 ax2.plot(y[5, 1:j], y[3, 1:j] * 1E2, '.-', linewidth=2, markersize=8, label=f.stem.replace('study_coco_', '').replace('yolo', 'YOLO')) ax2.plot(1E3 / np.array([209, 140, 97, 58, 35, 18]), [34.6, 40.5, 43.0, 47.5, 49.7, 51.5], 'k.-', linewidth=2, markersize=8, alpha=.25, label='EfficientDet') ax2.grid(alpha=0.2) ax2.set_yticks(np.arange(20, 60, 5)) ax2.set_xlim(0, 57) ax2.set_ylim(25, 55) ax2.set_xlabel('GPU Speed (ms/img)') ax2.set_ylabel('COCO AP val') ax2.legend(loc='lower right') f = save_dir / 'study.png' print(f'Saving {f}...') plt.savefig(f, dpi=300) @try_except # known issue https://github.com/ultralytics/yolov5/issues/5395 @Timeout(30) # known issue https://github.com/ultralytics/yolov5/issues/5611 def plot_labels(labels, names=(), save_dir=Path('')): # plot dataset labels LOGGER.info(f"Plotting labels to {save_dir / 'labels.jpg'}... ") c, b = labels[:, 0], labels[:, 1:].transpose() # classes, boxes nc = int(c.max() + 1) # number of classes x = pd.DataFrame(b.transpose(), columns=['x', 'y', 'width', 'height']) # seaborn correlogram sn.pairplot(x, corner=True, diag_kind='auto', kind='hist', diag_kws=dict(bins=50), plot_kws=dict(pmax=0.9)) plt.savefig(save_dir / 'labels_correlogram.jpg', dpi=200) plt.close() # matplotlib labels matplotlib.use('svg') # faster ax = plt.subplots(2, 2, figsize=(8, 8), tight_layout=True)[1].ravel() y = ax[0].hist(c, bins=np.linspace(0, nc, nc + 1) - 0.5, rwidth=0.8) # [y[2].patches[i].set_color([x / 255 for x in colors(i)]) for i in range(nc)] # update colors bug #3195 ax[0].set_ylabel('instances') if 0 < len(names) < 30: ax[0].set_xticks(range(len(names))) ax[0].set_xticklabels(names, rotation=90, fontsize=10) else: ax[0].set_xlabel('classes') sn.histplot(x, x='x', y='y', ax=ax[2], bins=50, pmax=0.9) sn.histplot(x, x='width', y='height', ax=ax[3], bins=50, pmax=0.9) # rectangles labels[:, 1:3] = 0.5 # center labels[:, 1:] = xywh2xyxy(labels[:, 1:]) * 2000 img = Image.fromarray(np.ones((2000, 2000, 3), dtype=np.uint8) * 255) for cls, *box in labels[:1000]: ImageDraw.Draw(img).rectangle(box, width=1, outline=colors(cls)) # plot ax[1].imshow(img) ax[1].axis('off') for a in [0, 1, 2, 3]: for s in ['top', 'right', 'left', 'bottom']: ax[a].spines[s].set_visible(False) plt.savefig(save_dir / 'labels.jpg', dpi=200) matplotlib.use('Agg') plt.close() def plot_evolve(evolve_csv='path/to/evolve.csv'): # from utils.plots import *; plot_evolve() # Plot evolve.csv hyp evolution results evolve_csv = Path(evolve_csv) data = pd.read_csv(evolve_csv) keys = [x.strip() for x in data.columns] x = data.values f = fitness(x) j = np.argmax(f) # max fitness index plt.figure(figsize=(10, 12), tight_layout=True) matplotlib.rc('font', **{'size': 8}) for i, k in enumerate(keys[7:]): v = x[:, 7 + i] mu = v[j] # best single result plt.subplot(6, 5, i + 1) plt.scatter(v, f, c=hist2d(v, f, 20), cmap='viridis', alpha=.8, edgecolors='none') plt.plot(mu, f.max(), 'k+', markersize=15) plt.title(f'{k} = {mu:.3g}', fontdict={'size': 9}) # limit to 40 characters if i % 5 != 0: plt.yticks([]) print(f'{k:>15}: {mu:.3g}') f = evolve_csv.with_suffix('.png') # filename plt.savefig(f, dpi=200) plt.close() print(f'Saved {f}') def plot_results(file='path/to/results.csv', dir=''): # Plot training results.csv. Usage: from utils.plots import *; plot_results('path/to/results.csv') save_dir = Path(file).parent if file else Path(dir) fig, ax = plt.subplots(2, 5, figsize=(12, 6), tight_layout=True) ax = ax.ravel() files = list(save_dir.glob('results*.csv')) assert len(files), f'No results.csv files found in {save_dir.resolve()}, nothing to plot.' for fi, f in enumerate(files): try: data = pd.read_csv(f) s = [x.strip() for x in data.columns] x = data.values[:, 0] for i, j in enumerate([1, 2, 3, 4, 5, 8, 9, 10, 6, 7]): y = data.values[:, j] # y[y == 0] = np.nan # don't show zero values ax[i].plot(x, y, marker='.', label=f.stem, linewidth=2, markersize=8) ax[i].set_title(s[j], fontsize=12) # if j in [8, 9, 10]: # share train and val loss y axes # ax[i].get_shared_y_axes().join(ax[i], ax[i - 5]) except Exception as e: print(f'Warning: Plotting error for {f}: {e}') ax[1].legend() fig.savefig(save_dir / 'results.png', dpi=200) plt.close() def profile_idetection(start=0, stop=0, labels=(), save_dir=''): # Plot iDetection '*.txt' per-image logs. from utils.plots import *; profile_idetection() ax = plt.subplots(2, 4, figsize=(12, 6), tight_layout=True)[1].ravel() s = ['Images', 'Free Storage (GB)', 'RAM Usage (GB)', 'Battery', 'dt_raw (ms)', 'dt_smooth (ms)', 'real-world FPS'] files = list(Path(save_dir).glob('frames*.txt')) for fi, f in enumerate(files): try: results = np.loadtxt(f, ndmin=2).T[:, 90:-30] # clip first and last rows n = results.shape[1] # number of rows x = np.arange(start, min(stop, n) if stop else n) results = results[:, x] t = (results[0] - results[0].min()) # set t0=0s results[0] = x for i, a in enumerate(ax): if i < len(results): label = labels[fi] if len(labels) else f.stem.replace('frames_', '') a.plot(t, results[i], marker='.', label=label, linewidth=1, markersize=5) a.set_title(s[i]) a.set_xlabel('time (s)') # if fi == len(files) - 1: # a.set_ylim(bottom=0) for side in ['top', 'right']: a.spines[side].set_visible(False) else: a.remove() except Exception as e: print(f'Warning: Plotting error for {f}; {e}') ax[1].legend() plt.savefig(Path(save_dir) / 'idetection_profile.png', dpi=200) def save_one_box(xyxy, im, file='image.jpg', gain=1.02, pad=10, square=False, BGR=False, save=True): # Save image crop as {file} with crop size multiple {gain} and {pad} pixels. Save and/or return crop xyxy = torch.tensor(xyxy).view(-1, 4) b = xyxy2xywh(xyxy) # boxes if square: b[:, 2:] = b[:, 2:].max(1)[0].unsqueeze(1) # attempt rectangle to square b[:, 2:] = b[:, 2:] * gain + pad # box wh * gain + pad xyxy = xywh2xyxy(b).long() clip_coords(xyxy, im.shape) crop = im[int(xyxy[0, 1]):int(xyxy[0, 3]), int(xyxy[0, 0]):int(xyxy[0, 2]), ::(1 if BGR else -1)] if save: file.parent.mkdir(parents=True, exist_ok=True) # make directory cv2.imwrite(str(increment_path(file).with_suffix('.jpg')), crop) return crop ================================================ FILE: utils/torch_utils.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ PyTorch utils """ import datetime import math import os import platform import subprocess import time from contextlib import contextmanager from copy import deepcopy from pathlib import Path import torch import torch.distributed as dist import torch.nn as nn import torch.nn.functional as F from utils.general import LOGGER try: import thop # for FLOPs computation except ImportError: thop = None @contextmanager def torch_distributed_zero_first(local_rank: int): """ Decorator to make all processes in distributed training wait for each local_master to do something. """ if local_rank not in [-1, 0]: dist.barrier(device_ids=[local_rank]) yield if local_rank == 0: dist.barrier(device_ids=[0]) def date_modified(path=__file__): # return human-readable file modification date, i.e. '2021-3-26' t = datetime.datetime.fromtimestamp(Path(path).stat().st_mtime) return f'{t.year}-{t.month}-{t.day}' def git_describe(path=Path(__file__).parent): # path must be a directory # return human-readable git description, i.e. v5.0-5-g3e25f1e https://git-scm.com/docs/git-describe s = f'git -C {path} describe --tags --long --always' try: return subprocess.check_output(s, shell=True, stderr=subprocess.STDOUT).decode()[:-1] except subprocess.CalledProcessError as e: return '' # not a git repository def select_device(device='', batch_size=0, newline=True): # device = 'cpu' or '0' or '0,1,2,3' s = f'YOLOv5 🚀 {git_describe() or date_modified()} torch {torch.__version__} ' # string device = str(device).strip().lower().replace('cuda:', '') # to string, 'cuda:0' to '0' cpu = device == 'cpu' if cpu: os.environ['CUDA_VISIBLE_DEVICES'] = '-1' # force torch.cuda.is_available() = False elif device: # non-cpu device requested os.environ['CUDA_VISIBLE_DEVICES'] = device # set environment variable assert torch.cuda.is_available(), f'CUDA unavailable, invalid device {device} requested' # check availability cuda = not cpu and torch.cuda.is_available() if cuda: devices = device.split(',') if device else '0' # range(torch.cuda.device_count()) # i.e. 0,1,6,7 n = len(devices) # device count if n > 1 and batch_size > 0: # check batch_size is divisible by device_count assert batch_size % n == 0, f'batch-size {batch_size} not multiple of GPU count {n}' space = ' ' * (len(s) + 1) for i, d in enumerate(devices): p = torch.cuda.get_device_properties(i) s += f"{'' if i == 0 else space}CUDA:{d} ({p.name}, {p.total_memory / 1024 ** 2:.0f}MiB)\n" # bytes to MB else: s += 'CPU\n' if not newline: s = s.rstrip() LOGGER.info(s.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else s) # emoji-safe return torch.device('cuda:0' if cuda else 'cpu') def time_sync(): # pytorch-accurate time if torch.cuda.is_available(): torch.cuda.synchronize() return time.time() def profile(input, ops, n=10, device=None): # YOLOv5 speed/memory/FLOPs profiler # # Usage: # input = torch.randn(16, 3, 640, 640) # m1 = lambda x: x * torch.sigmoid(x) # m2 = nn.SiLU() # profile(input, [m1, m2], n=100) # profile over 100 iterations results = [] device = device or select_device() print(f"{'Params':>12s}{'GFLOPs':>12s}{'GPU_mem (GB)':>14s}{'forward (ms)':>14s}{'backward (ms)':>14s}" f"{'input':>24s}{'output':>24s}") for x in input if isinstance(input, list) else [input]: x = x.to(device) x.requires_grad = True for m in ops if isinstance(ops, list) else [ops]: m = m.to(device) if hasattr(m, 'to') else m # device m = m.half() if hasattr(m, 'half') and isinstance(x, torch.Tensor) and x.dtype is torch.float16 else m tf, tb, t = 0, 0, [0, 0, 0] # dt forward, backward try: flops = thop.profile(m, inputs=(x,), verbose=False)[0] / 1E9 * 2 # GFLOPs except: flops = 0 try: for _ in range(n): t[0] = time_sync() y = m(x) t[1] = time_sync() try: _ = (sum(yi.sum() for yi in y) if isinstance(y, list) else y).sum().backward() t[2] = time_sync() except Exception as e: # no backward method # print(e) # for debug t[2] = float('nan') tf += (t[1] - t[0]) * 1000 / n # ms per op forward tb += (t[2] - t[1]) * 1000 / n # ms per op backward mem = torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0 # (GB) s_in = tuple(x.shape) if isinstance(x, torch.Tensor) else 'list' s_out = tuple(y.shape) if isinstance(y, torch.Tensor) else 'list' p = sum(list(x.numel() for x in m.parameters())) if isinstance(m, nn.Module) else 0 # parameters print(f'{p:12}{flops:12.4g}{mem:>14.3f}{tf:14.4g}{tb:14.4g}{str(s_in):>24s}{str(s_out):>24s}') results.append([p, flops, mem, tf, tb, s_in, s_out]) except Exception as e: print(e) results.append(None) torch.cuda.empty_cache() return results def is_parallel(model): # Returns True if model is of type DP or DDP return type(model) in (nn.parallel.DataParallel, nn.parallel.DistributedDataParallel) def de_parallel(model): # De-parallelize a model: returns single-GPU model if model is of type DP or DDP return model.module if is_parallel(model) else model def initialize_weights(model): for m in model.modules(): t = type(m) if t is nn.Conv2d: pass # nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu') elif t is nn.BatchNorm2d: m.eps = 1e-3 m.momentum = 0.03 elif t in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU]: m.inplace = True def find_modules(model, mclass=nn.Conv2d): # Finds layer indices matching module class 'mclass' return [i for i, m in enumerate(model.module_list) if isinstance(m, mclass)] def sparsity(model): # Return global model sparsity a, b = 0, 0 for p in model.parameters(): a += p.numel() b += (p == 0).sum() return b / a def prune(model, amount=0.3): # Prune model to requested global sparsity import torch.nn.utils.prune as prune print('Pruning model... ', end='') for name, m in model.named_modules(): if isinstance(m, nn.Conv2d): prune.l1_unstructured(m, name='weight', amount=amount) # prune prune.remove(m, 'weight') # make permanent print(' %.3g global sparsity' % sparsity(model)) def fuse_conv_and_bn(conv, bn): # Fuse convolution and batchnorm layers https://tehnokv.com/posts/fusing-batchnorm-and-conv/ fusedconv = nn.Conv2d(conv.in_channels, conv.out_channels, kernel_size=conv.kernel_size, stride=conv.stride, padding=conv.padding, groups=conv.groups, bias=True).requires_grad_(False).to(conv.weight.device) # prepare filters w_conv = conv.weight.clone().view(conv.out_channels, -1) w_bn = torch.diag(bn.weight.div(torch.sqrt(bn.eps + bn.running_var))) fusedconv.weight.copy_(torch.mm(w_bn, w_conv).view(fusedconv.weight.shape)) # prepare spatial bias b_conv = torch.zeros(conv.weight.size(0), device=conv.weight.device) if conv.bias is None else conv.bias b_bn = bn.bias - bn.weight.mul(bn.running_mean).div(torch.sqrt(bn.running_var + bn.eps)) fusedconv.bias.copy_(torch.mm(w_bn, b_conv.reshape(-1, 1)).reshape(-1) + b_bn) return fusedconv def model_info(model, verbose=False, img_size=640): # Model information. img_size may be int or list, i.e. img_size=640 or img_size=[640, 320] n_p = sum(x.numel() for x in model.parameters()) # number parameters n_g = sum(x.numel() for x in model.parameters() if x.requires_grad) # number gradients if verbose: print(f"{'layer':>5} {'name':>40} {'gradient':>9} {'parameters':>12} {'shape':>20} {'mu':>10} {'sigma':>10}") for i, (name, p) in enumerate(model.named_parameters()): name = name.replace('module_list.', '') print('%5g %40s %9s %12g %20s %10.3g %10.3g' % (i, name, p.requires_grad, p.numel(), list(p.shape), p.mean(), p.std())) try: # FLOPs from thop import profile stride = max(int(model.stride.max()), 32) if hasattr(model, 'stride') else 32 img = torch.zeros((1, model.yaml.get('ch', 3), stride, stride), device=next(model.parameters()).device) # input flops = profile(deepcopy(model), inputs=(img,), verbose=False)[0] / 1E9 * 2 # stride GFLOPs img_size = img_size if isinstance(img_size, list) else [img_size, img_size] # expand if int/float fs = ', %.1f GFLOPs' % (flops * img_size[0] / stride * img_size[1] / stride) # 640x640 GFLOPs except (ImportError, Exception): fs = '' LOGGER.info(f"Model Summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}") def scale_img(img, ratio=1.0, same_shape=False, gs=32): # img(16,3,256,416) # scales img(bs,3,y,x) by ratio constrained to gs-multiple if ratio == 1.0: return img else: h, w = img.shape[2:] s = (int(h * ratio), int(w * ratio)) # new size img = F.interpolate(img, size=s, mode='bilinear', align_corners=False) # resize if not same_shape: # pad/crop img h, w = (math.ceil(x * ratio / gs) * gs for x in (h, w)) return F.pad(img, [0, w - s[1], 0, h - s[0]], value=0.447) # value = imagenet mean def copy_attr(a, b, include=(), exclude=()): # Copy attributes from b to a, options to only include [...] and to exclude [...] for k, v in b.__dict__.items(): if (len(include) and k not in include) or k.startswith('_') or k in exclude: continue else: setattr(a, k, v) class EarlyStopping: # YOLOv5 simple early stopper def __init__(self, patience=30): self.best_fitness = 0.0 # i.e. mAP self.best_epoch = 0 self.patience = patience or float('inf') # epochs to wait after fitness stops improving to stop self.possible_stop = False # possible stop may occur next epoch def __call__(self, epoch, fitness): if fitness >= self.best_fitness: # >= 0 to allow for early zero-fitness stage of training self.best_epoch = epoch self.best_fitness = fitness delta = epoch - self.best_epoch # epochs without improvement self.possible_stop = delta >= (self.patience - 1) # possible stop may occur next epoch stop = delta >= self.patience # stop training if patience exceeded if stop: LOGGER.info(f'Stopping training early as no improvement observed in last {self.patience} epochs. ' f'Best results observed at epoch {self.best_epoch}, best model saved as best.pt.\n' f'To update EarlyStopping(patience={self.patience}) pass a new patience value, ' f'i.e. `python train.py --patience 300` or use `--patience 0` to disable EarlyStopping.') return stop class ModelEMA: """ Model Exponential Moving Average from https://github.com/rwightman/pytorch-image-models Keep a moving average of everything in the model state_dict (parameters and buffers). This is intended to allow functionality like https://www.tensorflow.org/api_docs/python/tf/train/ExponentialMovingAverage A smoothed version of the weights is necessary for some training schemes to perform well. This class is sensitive where it is initialized in the sequence of model init, GPU assignment and distributed training wrappers. """ def __init__(self, model, decay=0.9999, updates=0): # Create EMA self.ema = deepcopy(model.module if is_parallel(model) else model).eval() # FP32 EMA # if next(model.parameters()).device.type != 'cpu': # self.ema.half() # FP16 EMA self.updates = updates # number of EMA updates self.decay = lambda x: decay * (1 - math.exp(-x / 2000)) # decay exponential ramp (to help early epochs) for p in self.ema.parameters(): p.requires_grad_(False) def update(self, model): # Update EMA parameters with torch.no_grad(): self.updates += 1 d = self.decay(self.updates) msd = model.module.state_dict() if is_parallel(model) else model.state_dict() # model state_dict for k, v in self.ema.state_dict().items(): if v.dtype.is_floating_point: v *= d v += (1 - d) * msd[k].detach() def update_attr(self, model, include=(), exclude=('process_group', 'reducer')): # Update EMA attributes copy_attr(self.ema, model, include, exclude) ================================================ FILE: val.py ================================================ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Validate a trained YOLOv5 model accuracy on a custom dataset Usage: $ python path/to/val.py --data coco128.yaml --weights yolov5s.pt --img 640 """ import argparse import json import os import sys from pathlib import Path from threading import Thread import numpy as np import torch from tqdm import tqdm FILE = Path(__file__).resolve() ROOT = FILE.parents[0] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.common import DetectMultiBackend from utils.callbacks import Callbacks from utils.datasets import create_dataloader from utils.general import (LOGGER, box_iou, check_dataset, check_img_size, check_requirements, check_yaml, coco80_to_coco91_class, colorstr, increment_path, non_max_suppression, print_args, scale_coords, xywh2xyxy, xyxy2xywh) from utils.metrics import ConfusionMatrix, ap_per_class from utils.plots import output_to_target, plot_images, plot_val_study from utils.torch_utils import select_device, time_sync def save_one_txt(predn, save_conf, shape, file): # Save one txt result gn = torch.tensor(shape)[[1, 0, 1, 0]] # normalization gain whwh for *xyxy, conf, cls in predn.tolist(): xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh line = (cls, *xywh, conf) if save_conf else (cls, *xywh) # label format with open(file, 'a') as f: f.write(('%g ' * len(line)).rstrip() % line + '\n') def save_one_json(predn, jdict, path, class_map): # Save one JSON result {"image_id": 42, "category_id": 18, "bbox": [258.15, 41.29, 348.26, 243.78], "score": 0.236} image_id = int(path.stem) if path.stem.isnumeric() else path.stem box = xyxy2xywh(predn[:, :4]) # xywh box[:, :2] -= box[:, 2:] / 2 # xy center to top-left corner for p, b in zip(predn.tolist(), box.tolist()): jdict.append({'image_id': image_id, 'category_id': class_map[int(p[5])], 'bbox': [round(x, 3) for x in b], 'score': round(p[4], 5)}) def process_batch(detections, labels, iouv): """ Return correct predictions matrix. Both sets of boxes are in (x1, y1, x2, y2) format. Arguments: detections (Array[N, 6]), x1, y1, x2, y2, conf, class labels (Array[M, 5]), class, x1, y1, x2, y2 Returns: correct (Array[N, 10]), for 10 IoU levels """ correct = torch.zeros(detections.shape[0], iouv.shape[0], dtype=torch.bool, device=iouv.device) iou = box_iou(labels[:, 1:], detections[:, :4]) x = torch.where((iou >= iouv[0]) & (labels[:, 0:1] == detections[:, 5])) # IoU above threshold and classes match if x[0].shape[0]: matches = torch.cat((torch.stack(x, 1), iou[x[0], x[1]][:, None]), 1).cpu().numpy() # [label, detection, iou] if x[0].shape[0] > 1: matches = matches[matches[:, 2].argsort()[::-1]] matches = matches[np.unique(matches[:, 1], return_index=True)[1]] # matches = matches[matches[:, 2].argsort()[::-1]] matches = matches[np.unique(matches[:, 0], return_index=True)[1]] matches = torch.Tensor(matches).to(iouv.device) correct[matches[:, 1].long()] = matches[:, 2:3] >= iouv return correct @torch.no_grad() def run(data, weights=None, # model.pt path(s) batch_size=32, # batch size imgsz=640, # inference size (pixels) conf_thres=0.001, # confidence threshold iou_thres=0.6, # NMS IoU threshold task='val', # train, val, test, speed or study device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu workers=8, # max dataloader workers (per RANK in DDP mode) single_cls=False, # treat as single-class dataset augment=False, # augmented inference verbose=False, # verbose output save_txt=False, # save results to *.txt save_hybrid=False, # save label+prediction hybrid results to *.txt save_conf=False, # save confidences in --save-txt labels save_json=False, # save a COCO-JSON results file project=ROOT / 'runs/val', # save to project/name name='exp', # save to project/name exist_ok=False, # existing project/name ok, do not increment half=True, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference model=None, dataloader=None, save_dir=Path(''), plots=True, callbacks=Callbacks(), compute_loss=None, ): # Initialize/load model and set device training = model is not None if training: # called by train.py device, pt, jit, engine = next(model.parameters()).device, True, False, False # get model device, PyTorch model half &= device.type != 'cpu' # half precision only supported on CUDA model.half() if half else model.float() else: # called directly device = select_device(device, batch_size=batch_size) # Directories save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir # Load model model = DetectMultiBackend(weights, device=device, dnn=dnn) stride, pt, jit, engine = model.stride, model.pt, model.jit, model.engine imgsz = check_img_size(imgsz, s=stride) # check image size half &= (pt or jit or engine) and device.type != 'cpu' # half precision only supported by PyTorch on CUDA if pt or jit: model.model.half() if half else model.model.float() elif engine: batch_size = model.batch_size else: half = False batch_size = 1 # export.py models default to batch-size 1 device = torch.device('cpu') LOGGER.info(f'Forcing --batch-size 1 square inference shape(1,3,{imgsz},{imgsz}) for non-PyTorch backends') # Data data = check_dataset(data) # check # Configure model.eval() is_coco = isinstance(data.get('val'), str) and data['val'].endswith('coco/val2017.txt') # COCO dataset nc = 1 if single_cls else int(data['nc']) # number of classes iouv = torch.linspace(0.5, 0.95, 10).to(device) # iou vector for mAP@0.5:0.95 niou = iouv.numel() # Dataloader if not training: model.warmup(imgsz=(1, 3, imgsz, imgsz), half=half) # warmup pad = 0.0 if task == 'speed' else 0.5 task = task if task in ('train', 'val', 'test') else 'val' # path to train/val/test images dataloader = create_dataloader(data[task], imgsz, batch_size, stride, single_cls, pad=pad, rect=pt, workers=workers, prefix=colorstr(f'{task}: '))[0] seen = 0 confusion_matrix = ConfusionMatrix(nc=nc) names = {k: v for k, v in enumerate(model.names if hasattr(model, 'names') else model.module.names)} class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) s = ('%20s' + '%11s' * 6) % ('Class', 'Images', 'Labels', 'P', 'R', 'mAP@.5', 'mAP@.5:.95') dt, p, r, f1, mp, mr, map50, map = [0.0, 0.0, 0.0], 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 loss = torch.zeros(3, device=device) jdict, stats, ap, ap_class = [], [], [], [] pbar = tqdm(dataloader, desc=s, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar for batch_i, (im, targets, paths, shapes) in enumerate(pbar): t1 = time_sync() if pt or jit or engine: im = im.to(device, non_blocking=True) targets = targets.to(device) im = im.half() if half else im.float() # uint8 to fp16/32 im /= 255 # 0 - 255 to 0.0 - 1.0 nb, _, height, width = im.shape # batch size, channels, height, width t2 = time_sync() dt[0] += t2 - t1 # Inference out, train_out = model(im) if training else model(im, augment=augment, val=True) # inference, loss outputs dt[1] += time_sync() - t2 # Loss if compute_loss: loss += compute_loss([x.float() for x in train_out], targets)[1] # box, obj, cls # NMS targets[:, 2:] *= torch.Tensor([width, height, width, height]).to(device) # to pixels lb = [targets[targets[:, 0] == i, 1:] for i in range(nb)] if save_hybrid else [] # for autolabelling t3 = time_sync() out = non_max_suppression(out, conf_thres, iou_thres, labels=lb, multi_label=True, agnostic=single_cls) dt[2] += time_sync() - t3 # Metrics for si, pred in enumerate(out): labels = targets[targets[:, 0] == si, 1:] nl = len(labels) tcls = labels[:, 0].tolist() if nl else [] # target class path, shape = Path(paths[si]), shapes[si][0] seen += 1 if len(pred) == 0: if nl: stats.append((torch.zeros(0, niou, dtype=torch.bool), torch.Tensor(), torch.Tensor(), tcls)) continue # Predictions if single_cls: pred[:, 5] = 0 predn = pred.clone() scale_coords(im[si].shape[1:], predn[:, :4], shape, shapes[si][1]) # native-space pred # Evaluate if nl: tbox = xywh2xyxy(labels[:, 1:5]) # target boxes scale_coords(im[si].shape[1:], tbox, shape, shapes[si][1]) # native-space labels labelsn = torch.cat((labels[:, 0:1], tbox), 1) # native-space labels correct = process_batch(predn, labelsn, iouv) if plots: confusion_matrix.process_batch(predn, labelsn) else: correct = torch.zeros(pred.shape[0], niou, dtype=torch.bool) stats.append((correct.cpu(), pred[:, 4].cpu(), pred[:, 5].cpu(), tcls)) # (correct, conf, pcls, tcls) # Save/log if save_txt: save_one_txt(predn, save_conf, shape, file=save_dir / 'labels' / (path.stem + '.txt')) if save_json: save_one_json(predn, jdict, path, class_map) # append to COCO-JSON dictionary callbacks.run('on_val_image_end', pred, predn, path, names, im[si]) # Plot images if plots and batch_i < 3: f = save_dir / f'val_batch{batch_i}_labels.jpg' # labels Thread(target=plot_images, args=(im, targets, paths, f, names), daemon=True).start() f = save_dir / f'val_batch{batch_i}_pred.jpg' # predictions Thread(target=plot_images, args=(im, output_to_target(out), paths, f, names), daemon=True).start() # Compute metrics stats = [np.concatenate(x, 0) for x in zip(*stats)] # to numpy if len(stats) and stats[0].any(): tp, fp, p, r, f1, ap, ap_class = ap_per_class(*stats, plot=plots, save_dir=save_dir, names=names) ap50, ap = ap[:, 0], ap.mean(1) # AP@0.5, AP@0.5:0.95 mp, mr, map50, map = p.mean(), r.mean(), ap50.mean(), ap.mean() nt = np.bincount(stats[3].astype(np.int64), minlength=nc) # number of targets per class else: nt = torch.zeros(1) # Print results pf = '%20s' + '%11i' * 2 + '%11.3g' * 4 # print format LOGGER.info(pf % ('all', seen, nt.sum(), mp, mr, map50, map)) # Print results per class if (verbose or (nc < 50 and not training)) and nc > 1 and len(stats): for i, c in enumerate(ap_class): LOGGER.info(pf % (names[c], seen, nt[c], p[i], r[i], ap50[i], ap[i])) # Print speeds t = tuple(x / seen * 1E3 for x in dt) # speeds per image if not training: shape = (batch_size, 3, imgsz, imgsz) LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {shape}' % t) # Plots if plots: confusion_matrix.plot(save_dir=save_dir, names=list(names.values())) callbacks.run('on_val_end') # Save JSON if save_json and len(jdict): w = Path(weights[0] if isinstance(weights, list) else weights).stem if weights is not None else '' # weights anno_json = str(Path(data.get('path', '../coco')) / 'annotations/instances_val2017.json') # annotations json pred_json = str(save_dir / f"{w}_predictions.json") # predictions json LOGGER.info(f'\nEvaluating pycocotools mAP... saving {pred_json}...') with open(pred_json, 'w') as f: json.dump(jdict, f) try: # https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocoEvalDemo.ipynb check_requirements(['pycocotools']) from pycocotools.coco import COCO from pycocotools.cocoeval import COCOeval anno = COCO(anno_json) # init annotations api pred = anno.loadRes(pred_json) # init predictions api eval = COCOeval(anno, pred, 'bbox') if is_coco: eval.params.imgIds = [int(Path(x).stem) for x in dataloader.dataset.img_files] # image IDs to evaluate eval.evaluate() eval.accumulate() eval.summarize() map, map50 = eval.stats[:2] # update results (mAP@0.5:0.95, mAP@0.5) except Exception as e: LOGGER.info(f'pycocotools unable to run: {e}') # Return results model.float() # for training if not training: s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else '' LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}{s}") maps = np.zeros(nc) + map for i, c in enumerate(ap_class): maps[c] = ap[i] return (mp, mr, map50, map, *(loss.cpu() / len(dataloader)).tolist()), maps, t def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model.pt path(s)') parser.add_argument('--batch-size', type=int, default=32, help='batch size') parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='inference size (pixels)') parser.add_argument('--conf-thres', type=float, default=0.001, help='confidence threshold') parser.add_argument('--iou-thres', type=float, default=0.6, help='NMS IoU threshold') parser.add_argument('--task', default='val', help='train, val, test, speed or study') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--workers', type=int, default=8, help='max dataloader workers (per RANK in DDP mode)') parser.add_argument('--single-cls', action='store_true', help='treat as single-class dataset') parser.add_argument('--augment', action='store_true', help='augmented inference') parser.add_argument('--verbose', action='store_true', help='report mAP by class') parser.add_argument('--save-txt', action='store_true', help='save results to *.txt') parser.add_argument('--save-hybrid', action='store_true', help='save label+prediction hybrid results to *.txt') parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels') parser.add_argument('--save-json', action='store_true', help='save a COCO-JSON results file') parser.add_argument('--project', default=ROOT / 'runs/val', help='save to project/name') parser.add_argument('--name', default='exp', help='save to project/name') parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') opt = parser.parse_args() opt.data = check_yaml(opt.data) # check YAML opt.save_json |= opt.data.endswith('coco.yaml') opt.save_txt |= opt.save_hybrid print_args(FILE.stem, opt) return opt def main(opt): check_requirements(requirements=ROOT / 'requirements.txt', exclude=('tensorboard', 'thop')) if opt.task in ('train', 'val', 'test'): # run normally if opt.conf_thres > 0.001: # https://github.com/ultralytics/yolov5/issues/1466 LOGGER.info(f'WARNING: confidence threshold {opt.conf_thres} >> 0.001 will produce invalid mAP values.') run(**vars(opt)) else: weights = opt.weights if isinstance(opt.weights, list) else [opt.weights] opt.half = True # FP16 for fastest results if opt.task == 'speed': # speed benchmarks # python val.py --task speed --data coco.yaml --batch 1 --weights yolov5n.pt yolov5s.pt... opt.conf_thres, opt.iou_thres, opt.save_json = 0.25, 0.45, False for opt.weights in weights: run(**vars(opt), plots=False) elif opt.task == 'study': # speed vs mAP benchmarks # python val.py --task study --data coco.yaml --iou 0.7 --weights yolov5n.pt yolov5s.pt... for opt.weights in weights: f = f'study_{Path(opt.data).stem}_{Path(opt.weights).stem}.txt' # filename to save to x, y = list(range(256, 1536 + 128, 128)), [] # x axis (image sizes), y axis for opt.imgsz in x: # img-size LOGGER.info(f'\nRunning {f} --imgsz {opt.imgsz}...') r, _, t = run(**vars(opt), plots=False) y.append(r + t) # results and times np.savetxt(f, y, fmt='%10.4g') # save os.system('zip -r study.zip study_*.txt') plot_val_study(x=x) # plot if __name__ == "__main__": opt = parse_opt() main(opt)