clarifying register model

This commit is contained in:
Justin 2025-10-06 15:09:38 -04:00
parent 0ba4cd460f
commit b519434c88

View file

@ -93,10 +93,13 @@ class RunpodInferenceAdapter(OpenAIMixin):
async def register_model(self, model: Model) -> Model:
"""
Register a model and verify it's available on the RunPod endpoint.
In the .yaml file the model: can be defined as example
This is mainly if you want to register a model with a custom identifier.
This will ping the endpoint and make sure the model is avaliable via the /v1/models.
In the .yaml file the model: can be defined as example.
models:
- metadata: {}
model_id: qwen3-32b-awq
model_id: custom_model_id
model_type: llm
provider_id: runpod
provider_model_id: Qwen/Qwen3-32B-AWQ