知微
fa45d8aa5f
- health_checklist.json: 192.168.1.122→node122
- ocr_client.py: docstring IP→node122
- docs/market-data-requirements.md: IP→node122
- 所有API调用通过ProxyHandler({})绕过系统代理
Privoxy对node122:18003返回500，直连正常
No transformation is required for hosted_vllm embedding.
VLLM is a superset of OpenAI's embedding endpoint.
encoding_format
For OpenAI-compatible embedding calls (including openai/... with a custom api_base pointing at vLLM), LiteLLM resolves encoding_format when it is not set on the request:
- Explicit value on the embedding call (
encoding_format=...). - Model config (
litellm_params.encoding_formaton the proxymodel_listentry). - Environment variable
LITELLM_DEFAULT_EMBEDDING_ENCODING_FORMAT(e.g. in.envor container env). - Default
float.
That avoids forwarding encoding_format=None to the provider/SDK where some servers behave poorly.
To pass provider-specific parameters, see provider-specific params.