Add Dockerfile-conda FROM continuumio/miniconda3:latest (#4706)

.github/workflows/ci.yaml

@@ -144,6 +144,7 @@ jobs:
         run: |
           coverage xml -o coverage-benchmarks.xml
       - name: Upload Coverage Reports to CodeCov
+        if: github.repository == 'ultralytics/ultralytics'
         uses: codecov/codecov-action@v3
         with:
           flags: Benchmarks
@@ -235,13 +236,73 @@ jobs:
         env:
           CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
 
+  Conda:
+    if: github.repository == 'ultralytics/ultralytics' && (github.event_name == 'workflow_dispatch' || github.event_name == 'schedule')
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest]
+        python-version: ['3.11']
+    defaults:
+      run:
+        shell: bash -el {0}
+    steps:
+      - uses: conda-incubator/setup-miniconda@v2
+        with:
+          python-version: ${{ matrix.python-version }}
+          mamba-version: "*"
+          channels: conda-forge,defaults
+          channel-priority: true
+          activate-environment: anaconda-client-env
+      - name: Install Libmamba
+        run: |
+          # conda install conda-libmamba-solver
+          conda config --set solver libmamba
+      - name: Install Ultralytics package from conda-forge
+        run: |
+          conda install -c conda-forge ultralytics
+      - name: Install pip packages
+        run: |
+          pip install pytest 'coremltools>=7.0.b1' # 'openvino-dev>=2023.0'
+      - name: Check environment
+        run: |
+          echo "RUNNER_OS is ${{ runner.os }}"
+          echo "GITHUB_EVENT_NAME is ${{ github.event_name }}"
+          echo "GITHUB_WORKFLOW is ${{ github.workflow }}"
+          echo "GITHUB_ACTOR is ${{ github.actor }}"
+          echo "GITHUB_REPOSITORY is ${{ github.repository }}"
+          echo "GITHUB_REPOSITORY_OWNER is ${{ github.repository_owner }}"
+          python --version
+          conda list
+      - name: Test CLI
+        run: |
+          yolo predict model=yolov8n.pt imgsz=320
+          yolo train model=yolov8n.pt data=coco8.yaml epochs=1 imgsz=32
+          yolo val model=yolov8n.pt data=coco8.yaml imgsz=32
+          yolo export model=yolov8n.pt format=torchscript imgsz=160
+      - name: Test Python
+        run: |
+          python -c "
+          from ultralytics import YOLO
+          model = YOLO('yolov8n.pt')
+          results = model.train(data='coco8.yaml', epochs=3, imgsz=160)
+          results = model.val(imgsz=160)
+          results = model.predict(imgsz=160)
+          results = model.export(format='onnx', imgsz=160)
+          "
+      - name: PyTest
+        run: |
+          git clone https://github.com/ultralytics/ultralytics
+          pytest ultralytics/tests/test_cli.py  # full tests fail due to openvino export failure
+
   Summary:
     runs-on: ubuntu-latest
-    needs: [HUB, Benchmarks, Tests, GPU] # Add job names that you want to check for failure
+    needs: [HUB, Benchmarks, Tests, GPU, Conda] # Add job names that you want to check for failure
     if: always() # This ensures the job runs even if previous jobs fail
     steps:
       - name: Check for failure and notify
-        if: (needs.HUB.result == 'failure' || needs.Benchmarks.result == 'failure' || needs.Tests.result == 'failure' || needs.GPU.result == 'failure') && github.repository == 'ultralytics/ultralytics' && (github.event_name == 'schedule' || github.event_name == 'push')
+        if: (needs.HUB.result == 'failure' || needs.Benchmarks.result == 'failure' || needs.Tests.result == 'failure' || needs.GPU.result == 'failure' || needs.Conda.result == 'failure') && github.repository == 'ultralytics/ultralytics' && (github.event_name == 'schedule' || github.event_name == 'push')
         uses: slackapi/slack-github-action@v1.24.0
         with:
           payload: |

.github/workflows/docker.yaml

@@ -24,6 +24,9 @@ on:
       Dockerfile-python:
         type: boolean
         description: Use Dockerfile-python
+      Dockerfile-conda:
+        type: boolean
+        description: Use Dockerfile-conda
       push:
         type: boolean
         description: Push images to Docker Hub
@@ -36,9 +39,15 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       fail-fast: false
-      max-parallel: 5
+      max-parallel: 6
       matrix:
         include:
+          - dockerfile: "Dockerfile"
+            tags: "latest"
+            platforms: "linux/amd64"
+          - dockerfile: "Dockerfile-cpu"
+            tags: "latest-cpu"
+            platforms: "linux/amd64"
           - dockerfile: "Dockerfile-arm64"
             tags: "latest-arm64"
             platforms: "linux/arm64"
@@ -48,12 +57,9 @@ jobs:
           - dockerfile: "Dockerfile-python"
             tags: "latest-python"
             platforms: "linux/amd64"
-          - dockerfile: "Dockerfile-cpu"
+          # - dockerfile: "Dockerfile-conda"
-            tags: "latest-cpu"
+          #   tags: "latest-conda"
-            platforms: "linux/amd64"
+          #   platforms: "linux/amd64"
-          - dockerfile: "Dockerfile"
-            tags: "latest"
-            platforms: "linux/amd64"
     steps:
       - name: Checkout repo
         uses: actions/checkout@v3
@@ -109,16 +115,16 @@ jobs:
           -t ultralytics/ultralytics:${{ steps.get_version.outputs.version_tag }} .
 
       - name: Run Tests
-        if: (github.event_name == 'push' || github.event.inputs[matrix.dockerfile] == 'true') && matrix.platforms == 'linux/amd64'  # arm64 images not supported on GitHub CI runners
+        if: (github.event_name == 'push' || github.event.inputs[matrix.dockerfile] == 'true') && matrix.platforms == 'linux/amd64' && matrix.dockerfile != 'Dockerfile-conda'  # arm64 images not supported on GitHub CI runners
         run: docker run ultralytics/ultralytics:${{ matrix.tags }} /bin/bash -c "pip install pytest && pytest tests"
 
       - name: Run Benchmarks
         # WARNING: Dockerfile (GPU) error on TF.js export 'module 'numpy' has no attribute 'object'.
-        if: (github.event_name == 'push' || github.event.inputs[matrix.dockerfile] == 'true') && matrix.platforms == 'linux/amd64' && matrix.dockerfile != 'Dockerfile'  # arm64 images not supported on GitHub CI runners
+        if: (github.event_name == 'push' || github.event.inputs[matrix.dockerfile] == 'true') && matrix.platforms == 'linux/amd64' && matrix.dockerfile != 'Dockerfile' && matrix.dockerfile != 'Dockerfile-conda'  # arm64 images not supported on GitHub CI runners
         run: docker run ultralytics/ultralytics:${{ matrix.tags }} yolo benchmark model=yolov8n.pt imgsz=160 verbose=0.26
 
       - name: Push Docker Image with Ultralytics version tag
-        if: (github.event_name == 'push' || (github.event.inputs[matrix.dockerfile] == 'true' && github.event.inputs.push == 'true')) && steps.check_tag.outputs.exists == 'false'
+        if: (github.event_name == 'push' || (github.event.inputs[matrix.dockerfile] == 'true' && github.event.inputs.push == 'true')) && steps.check_tag.outputs.exists == 'false' && matrix.dockerfile != 'Dockerfile-conda'
         run: |
           docker push ultralytics/ultralytics:${{ steps.get_version.outputs.version_tag }}
 

docker/Dockerfile-conda

@@ -0,0 +1,37 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+# Builds ultralytics/ultralytics:latest-conda image on DockerHub https://hub.docker.com/r/ultralytics/ultralytics
+# Image is optimized for Ultralytics Anaconda (https://anaconda.org/conda-forge/ultralytics) installation and usage
+
+# Start FROM miniconda3 image https://hub.docker.com/r/continuumio/miniconda3
+FROM continuumio/miniconda3:latest
+
+# Downloads to user config dir
+ADD https://ultralytics.com/assets/Arial.ttf https://ultralytics.com/assets/Arial.Unicode.ttf /root/.config/Ultralytics/
+
+# Install linux packages
+RUN apt update \
+    && apt install --no-install-recommends -y libgl1-mesa-glx
+
+# Copy contents
+ADD https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8n.pt .
+
+# Install conda packages
+# mkl required to fix 'OSError: libmkl_intel_lp64.so.2: cannot open shared object file: No such file or directory'
+RUN conda config --set solver libmamba && \
+    conda install -c conda-forge ultralytics mkl
+    # conda install -c pytorch -c nvidia -c conda-forge pytorch torchvision pytorch-cuda=11.8 ultralytics
+
+
+# Usage Examples -------------------------------------------------------------------------------------------------------
+
+# Build and Push
+# t=ultralytics/ultralytics:latest-conda && sudo docker build -f docker/Dockerfile-cpu -t $t . && sudo docker push $t
+
+# Run
+# t=ultralytics/ultralytics:latest-conda && sudo docker run -it --ipc=host $t
+
+# Pull and Run
+# t=ultralytics/ultralytics:latest-conda && sudo docker pull $t && sudo docker run -it --ipc=host $t
+
+# Pull and Run with local volume mounted
+# t=ultralytics/ultralytics:latest-conda && sudo docker pull $t && sudo docker run -it --ipc=host -v "$(pwd)"/datasets:/usr/src/datasets $t

docker/Dockerfile-cpu

@@ -37,6 +37,7 @@ RUN yolo export model=tmp/yolov8n.pt format=ncnn imgsz=32
 # Remove exported models
 RUN rm -rf tmp
 
+
 # Usage Examples -------------------------------------------------------------------------------------------------------
 
 # Build and Push

docker/Dockerfile-python

@@ -37,6 +37,7 @@ RUN pip install --no-cache paddlepaddle==2.4.2 x2paddle
 # Remove exported models
 RUN rm -rf tmp
 
+
 # Usage Examples -------------------------------------------------------------------------------------------------------
 
 # Build and Push

docs/integrations/index.md

@@ -12,7 +12,7 @@ Welcome to the Ultralytics Integrations page! This page provides an overview of
 
 ## Datasets Integrations
 
-- [Roboflow](/integrations/roboflow/): Facilitate seamless dataset management for Ultralytics models, offering robust annotation, preprocessing, and augmentation capabilities.
+- [Roboflow](roboflow.md): Facilitate seamless dataset management for Ultralytics models, offering robust annotation, preprocessing, and augmentation capabilities.
 
 ## Training Integrations
 

docs/quickstart.md

@@ -20,7 +20,7 @@ Ultralytics provides various installation methods including pip, conda, and Dock
         pip install ultralytics
         ```
 
-        You can also install the `ultralytics` package directly from the GitHub repository. This might be useful if you want the latest development version. Make sure to have the Git command-line tool installed on your system. The `@main` command installs the `main` branch and may be modified to another branch, i.e. `@my-branch`, or removed alltogether to default to `main` branch.
+        You can also install the `ultralytics` package directly from the GitHub [repository](https://github.com/ultralytics/ultralytics). This might be useful if you want the latest development version. Make sure to have the Git command-line tool installed on your system. The `@main` command installs the `main` branch and may be modified to another branch, i.e. `@my-branch`, or removed alltogether to default to `main` branch.
     
         ```bash
         # Install the ultralytics package from GitHub
@@ -44,7 +44,23 @@ Ultralytics provides various installation methods including pip, conda, and Dock
             If you are installing in a CUDA environment best practice is to install `ultralytics`, `pytorch` and `pytorch-cuda` in the same command to allow the conda package manager to resolve any conflicts, or else to install `pytorch-cuda` last to allow it override the CPU-specific `pytorch` package if necessary.
             ```bash
             # Install all packages together using conda
-            conda install -c conda-forge -c pytorch -c nvidia ultralytics pytorch torchvision pytorch-cuda=11.8 
+            conda install -c pytorch -c nvidia -c conda-forge pytorch torchvision pytorch-cuda=11.8 ultralytics 
+            ```
+
+        ### Conda Docker Image
+    
+        Ultralytics Conda Docker images are also available from [DockerHub](https://hub.docker.com/r/ultralytics/ultralytics). These images are based on [Miniconda3](https://docs.conda.io/projects/miniconda/en/latest/) and are an simple way to start using `ultralytics` in a Conda environment.
+
+        ```bash
+        # Set image name as a variable
+        t=ultralytics/ultralytics:latest-conda
+
+        # Pull the latest ultralytics image from Docker Hub
+        sudo docker pull $t
+
+        # Run the ultralytics image in a container with GPU support
+        sudo docker run -it --ipc=host --gpus all $t  # all GPUs
+        sudo docker run -it --ipc=host --gpus '"device=2,3"' $t  # specify GPUs
         ```
 
     === "Git clone"
@@ -71,6 +87,7 @@ Ultralytics provides various installation methods including pip, conda, and Dock
         - **Dockerfile-cpu:** Ubuntu-based CPU-only version suitable for inference and environments without GPUs.
         - **Dockerfile-jetson:** Tailored for NVIDIA Jetson devices, integrating GPU support optimized for these platforms.
         - **Dockerfile-python:** Minimal image with just Python and necessary dependencies, ideal for lightweight applications and development.
+        - **Dockerfile-conda:** Based on Miniconda3 with conda installation of ultralytics package.
         
         Below are the commands to get the latest image and execute it:
 
@@ -109,8 +126,7 @@ See the `ultralytics` [requirements.txt](https://github.com/ultralytics/ultralyt
 
 ## Use Ultralytics with CLI
 
-The Ultralytics command line interface (CLI) allows for simple single-line commands without the need for a Python environment.
+The Ultralytics command line interface (CLI) allows for simple single-line commands without the need for a Python environment. CLI requires no customization or Python code. You can simply run all tasks from the terminal with the `yolo` command. Check out the [CLI Guide](usage/cli.md) to learn more about using YOLOv8 from the command line.
-CLI requires no customization or Python code. You can simply run all tasks from the terminal with the `yolo` command. Check out the [CLI Guide](usage/cli.md) to learn more about using YOLOv8 from the command line.
 
 !!! example
 

ultralytics/engine/exporter.py

@@ -154,7 +154,8 @@ class Exporter:
         format = self.args.format.lower()  # to lowercase
         if format in ('tensorrt', 'trt'):  # 'engine' aliases
             format = 'engine'
-        if format in ('mlmodel', 'mlpackage', 'mlprogram', 'apple', 'ios'):  # 'coreml' aliases
+        if format in ('mlmodel', 'mlpackage', 'mlprogram', 'apple', 'ios', 'coreml'):  # 'coreml' aliases
+            os.environ['PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION'] = 'python'  # fix attempt for protobuf<3.20.x errors
             format = 'coreml'
         fmts = tuple(export_formats()['Argument'][1:])  # available export formats
         flags = [x == format for x in fmts]

ultralytics/engine/validator.py

@@ -24,7 +24,6 @@ from pathlib import Path
 
 import numpy as np
 import torch
-from scipy.optimize import linear_sum_assignment
 
 from ultralytics.cfg import get_cfg, get_save_dir
 from ultralytics.data.utils import check_cls_dataset, check_det_dataset
@@ -226,9 +225,11 @@ class BaseValidator:
         iou = iou.cpu().numpy()
         for i, threshold in enumerate(self.iouv.cpu().tolist()):
             if use_scipy:
+                # WARNING: known issue that reduces mAP in https://github.com/ultralytics/ultralytics/pull/4708
+                import scipy  # scope import to avoid importing for all commands
                 cost_matrix = iou * (iou >= threshold)
                 if cost_matrix.any():
-                    labels_idx, detections_idx = linear_sum_assignment(cost_matrix, maximize=True)
+                    labels_idx, detections_idx = scipy.optimize.linear_sum_assignment(cost_matrix, maximize=True)
                     valid = cost_matrix[labels_idx, detections_idx] > 0
                     if valid.any():
                         correct[detections_idx[valid], i] = True

ultralytics/models/utils/ops.py

@@ -14,11 +14,11 @@ class HungarianMatcher(nn.Module):
     A module implementing the HungarianMatcher, which is a differentiable module to solve the assignment problem in
     an end-to-end fashion.
 
-    HungarianMatcher performs optimal assignment over predicted and ground truth bounding boxes using a cost function
+    HungarianMatcher performs optimal assignment over the predicted and ground truth bounding boxes using a cost
-    that considers classification scores, bounding box coordinates, and optionally, mask predictions.
+    function that considers classification scores, bounding box coordinates, and optionally, mask predictions.
 
     Attributes:
-        cost_gain (dict): Dictionary of cost coefficients for different components: 'class', 'bbox', 'giou', 'mask', and 'dice'.
+        cost_gain (dict): Dictionary of cost coefficients: 'class', 'bbox', 'giou', 'mask', and 'dice'.
         use_fl (bool): Indicates whether to use Focal Loss for the classification cost calculation.
         with_mask (bool): Indicates whether the model makes mask predictions.
         num_sample_points (int): The number of sample points used in mask cost calculation.
@@ -26,8 +26,8 @@ class HungarianMatcher(nn.Module):
         gamma (float): The gamma factor in Focal Loss calculation.
 
     Methods:
-        forward(pred_bboxes, pred_scores, gt_bboxes, gt_cls, gt_groups, masks=None, gt_mask=None): Computes the assignment
+        forward(pred_bboxes, pred_scores, gt_bboxes, gt_cls, gt_groups, masks=None, gt_mask=None): Computes the
-        between predictions and ground truths for a batch.
+            assignment between predictions and ground truths for a batch.
         _cost_mask(bs, num_gts, masks=None, gt_mask=None): Computes the mask cost and dice cost if masks are predicted.
     """