Skip to main content

Image Issues

Solutions for Docker image building and package problems.

Image Build Fails​

Symptoms:

$ image-create my-project
Error: Failed to build image

Common causes and solutions:

Network Issues​

# Retry build
image-create my-project

# Check if base image cached
docker images | grep aime-pytorch

Package Installation Fails​

# Use interactive GUI to fix packages
image-update                  # Select image, fix package names/versions

# Or check Dockerfile manually
cat ~/dockerfiles/my-project.Dockerfile
vim ~/dockerfiles/my-project.Dockerfile
image-update my-project --rebuild

Disk Space​

Best to notify DSL admin by raising an issue ticket in ds01-hub repo. Most user permissions are restricted so you will not be able to do a full clean of docker except from those files related to you (limited).

df -h
docker system df
docker system prune  # Free space

Invalid Dockerfile Syntax​

# Validate Dockerfile
docker build --no-cache -f ~/dockerfiles/my-project.Dockerfile . 2>&1 | less

Package Not Found​

Symptoms:

ModuleNotFoundError: No module named 'transformers'

Causes:

  • Package not in image
  • Package name typo

Note: DS01 containers ARE your Python environment - you don't need venv or conda. See Python Environments.

Solutions:

  1. Check if installed:

    pip list | grep transformers

  2. Temporary install:

    pip install transformers

  3. Permanent fix (add to image):

    exit  # Exit container
    image-update                  # Select image, add package
    container-retire <project-name>
    container-deploy <project-name>

Image Too Large​

Symptoms:

  • Build takes very long
  • "No space left on device"

Solutions:

  1. Check image size:

    docker images | grep my-project

  2. Use .dockerignore:

    echo "data/" >> ~/workspace/<my-project>/.dockerignore
    echo "*.csv" >> ~/workspace/<my-project>/.dockerignore
    echo "models/" >> ~/workspace/<my-project>/.dockerignore

  3. Combine RUN commands:

    # Bad (creates extra layers)
    RUN pip install package1
    RUN pip install package2

    # Good (single layer)
    RUN pip install package1 package2

  4. Clean up in same layer:

    RUN pip install --no-cache-dir packages && \
        apt-get clean && \
        rm -rf /var/lib/apt/lists/*

Image Won't Update​

Symptoms:

  • Changes to Dockerfile not reflected
  • Old packages still installed

Solutions:

  1. Rebuild without cache: 
    image-update my-project --no-cache
  2. Recreate container after rebuild: 
    container-deploy my-project

Base Image Not Found​

Symptoms:

Error: manifest for henrycgbaker/aime-pytorch:2.8.0-cuda12.4-ubuntu22.04 not found

Solutions:

  1. Check available base images:

    docker images | grep aime

  2. Pull base image:

    docker pull henrycgbaker/aime-pytorch:2.8.0-cuda12.4-ubuntu22.04

  3. Use different base image version:

    # Edit Dockerfile
    vim ~/workspace/my-project/Dockerfile
    # Change FROM line to available image

Dependency Conflicts​

Symptoms:

ERROR: Cannot install package-a and package-b because these package versions have conflicting dependencies

Solutions:

  1. Pin specific versions:

    RUN pip install transformers==4.30.0 datasets==2.14.0

  2. Install in order:

    RUN pip install torch==2.0.0 && \
        pip install transformers

  3. Create fresh environment:

    RUN pip install --upgrade pip && \
        pip install package1 package2

See Also​