canonical / canonical/inference-snaps

Requires two CPUs but doesn't mention that

Open
#90 0 comments 0 reactions 0 assignees View on GitHub
snap/qwen-vl
Dominant language
No language data
Stars
50
Forks
6
Avg merge
1d 20h
Merged PRs (30d)
7

Description

It seems to check the amount of RAM and disk needed, but not the number of CPUs. You get this error if you only have one:

```
2025-10-24T22:16:40Z qwen-vl.server[6696]: LD_LIBRARY_PATH: /var/lib/snapd/lib/gl:/var/lib/snapd/lib/gl32:/var/lib/snapd/void:/snap/qwen-vl/201/usr/lib:/snap/qwen-vl/201/usr/lib/x86_64-linux-gnu:/snap/qwen-vl/components/201/openvino-model-server/lib:/snap/qwen-vl/components/201/openvino-model-server/lib/python:/snap/qwen-vl/components/201/openvino-model-server/usr/lib/x86_64-linux-gnu:/snap/qwen-vl/201/usr/local/lib
2025-10-24T22:16:40Z qwen-vl.server[6696]: PYTHONPATH: /snap/qwen-vl/components/201/openvino-model-server/lib/python:/snap/qwen-vl/components/201/openvino-model-server/lib/python3.12/site-packages
2025-10-24T22:16:41Z qwen-vl.server[6696]: ++ /snap/qwen-vl/components/201/openvino-model-server/bin/ovms --rest_port 8326 --rest_bind_address 127.0.0.1 --source_model Qwen2.5-VL-3B-Instruct-ov-int4 --model_repository_path /snap/qwen-vl/components/201/model-qwen2-5-vl-3b-instruct-ov-int4 --target_device CPU --task text_generation --cache_size 2
2025-10-24T22:16:41Z qwen-vl.server[6824]: rest_workers count should be from 2 to 10000
```

Contributor guide

Open the contributing guide

Research direction

Start by reproducing the Qwen-VL server failure on a one-CPU system and inspect the resource checks and snap configuration in the inference-snaps repository. Done means the CPU requirement is checked or clearly reported alongside the existing RAM and disk requirements, with a test or documented reproduction covering the one-CPU case.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
ai, backend
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.