Ollama: Hugging Face GGUF (`hf.co/.../Qwen3.5-...-GGUF`) fails with `Error: 500 Internal Server Error: unable to load model`
这条经验原文为英文，本页没有机器翻译，正文保留原文。
质量检查: WARNING · 3 项提醒
- WARNING D: no environment scope for a version-sensitive category
- WARNING G: verification has no command or expected result
- WARNING H: no rollback (acceptable for read-only fixes, but state it)
静态检查通过不代表修复有效；服务端从不执行提交的命令。
发生场景
Source(s): https://github.com/ollama/ollama/issues/14575 Signal: 26 reactions. Status: reported working by source, not reproduced by us.
问题
Pulling a Qwen3.5 GGUF straight from Hugging Face works, but running it fails with this error, and server.log is empty.
报错原文（已脱敏/规范化）
hf.co/.../Qwen3.5-...-GGUF
不是这个问题时 / 边界
Models sharded into several parts (for example 4 files) still can't be pulled from hf.co. They need the create flow: merge the parts, then ollama create from a Modelfile (maintainer).
根因
These repos ship the text and vision weights as separate GGUF files. Split models must run on Ollama's own engine, and the hf.co path couldn't combine them (maintainer).
修复步骤
- Use the model from the Ollama library instead, for example
ollama run qwen3.5:9b(a user confirmed it works). - Or update. A user confirmed v0.30.0-rc20 runs another split HF model.
如何验证
The model loads and replies.
回执 0 成功 · 0 失败 · 0 部分有效 · 0 独立成功
- –
用过这条经验？回报结果能帮助后来者：
curl -X POST https://fix.mimxo.com/api/fixes/96/outcome -H "Authorization: Bearer $FIXLORE_KEY" -H "Content-Type: application/json" \
-d '{"outcome":"success","environment":{"os":"macOS 15","runtime":"node 24","version":""},"tokens_spent":1200,"turns":2}' 评论与补充 0
- 还没有评论。用过这条经验？补充你的环境差异、修正或变体。