Container to upscale images. Works as CLI, an HTTP server and an MCP tool. CUDA.
150
docker run --rm --gpus all \
-v "$(pwd)/upscaler/tests/fixtures:/data" \
sinanozel/upscaler:upscaler-realesrgan \
upscale --file /data/sample.png --scale 4 --output /data/sample.x4.png
docker run -d --gpus all -p 8080:8080 sinanozel/upscaler:upscaler-realesrgan
# Submit
curl -F "file=@upscaler/tests/fixtures/sample.png" -F "scale=4" \
http://localhost:8080/v1/upscale
# -> {"job_id": "8f14e45f-ceea-4e94-b4c3-8ddd9e2b0e18", "status": "pending"}
# Poll
curl http://localhost:8080/v1/upscale/8f14e45f-ceea-4e94-b4c3-8ddd9e2b0e18
# -> {"job_id": "...", "status": "completed", "progress": 1.0, "error": null}
# Download
curl http://localhost:8080/v1/upscale/8f14e45f-ceea-4e94-b4c3-8ddd9e2b0e18/result \
-o sample.x4.png
More in here, and you can also find the build+test pipeline: https://github.com/sinan-ozel/model-servers/blob/main/upscaler/README.md
Content type
Image
Digest
sha256:47df94d68…
Size
3.6 GB
Last updated
10 days ago
docker pull sinanozel/upscaler:realesrgan-cuda