mirror of
https://github.com/containers/ramalama.git
synced 2026-02-05 15:47:26 +01:00
630b75ec55
This commit introduces the `--generate=compose` option to the `ramalama serve` command, enabling users to generate a `docker-compose.yaml` file for a given model. sourcery-ai suggested changes: 1. Add test_genfile_empty_content. This will verify that genfile handles empty input gracefully and avoids unexpected behavior. 2. Supporting condition to handle 'oci://' prefix for RAG source for consistency. 3. updating the parsing logic to support complex port formats, such as those i with IP addresses or protocols, to ensure compatibility with all valid Docker Compose specifications. 4. approach excludes images for other GPU backends like ROCm or custom builds. Please update the check to support a wider range of GPU-enabled images. fixes #184 Co-authored-by: sourcery-ai[bot] <58596630+sourcery-ai[bot]@users.noreply.github.com> Signed-off-by: abhibongale <abhishekbongale@outlook.com>
20 lines
554 B
YAML
20 lines
554 B
YAML
# Save this output to a 'docker-compose.yaml' file and run 'docker compose up'.
|
|
#
|
|
# Created with ramalama-0.1.0-test
|
|
services:
|
|
tinyllama:
|
|
container_name: ramalama-tinyllama
|
|
image: test-image/ramalama:latest
|
|
volumes:
|
|
- "/models/tinyllama.gguf:/mnt/models/tinyllama.gguf:ro"
|
|
ports:
|
|
- "8080:8080"
|
|
environment:
|
|
- ACCEL_ENV=true
|
|
devices:
|
|
- "/dev/dri:/dev/dri"
|
|
- "/dev/kfd:/dev/kfd"
|
|
- "/dev/accel:/dev/accel"
|
|
command: llama-server --model /mnt/models/tinyllama.gguf
|
|
restart: unless-stopped
|