microsoft/autogen · error · ValueError

Failed to create custom embedding function. Error: {e}

Error message

Failed to create custom embedding function. Error: {e}

What it means

CustomEmbeddingFunctionConfig lets you supply an arbitrary callable plus params; it is invoked as config.function(**config.params) inside a try/except, and any exception it raises is re-raised as ValueError with the original error chained. This reports failures in user-supplied embedding factory code.

Source

Thrown at python/packages/autogen-ext/src/autogen_ext/memory/chromadb/_chromadb.py:233

                raise ImportError(
                    f"Failed to create SentenceTransformer embedding function with model '{config.model_name}'. "
                    f"Ensure sentence-transformers is installed and the model is available. Error: {e}"
                ) from e

        elif isinstance(config, OpenAIEmbeddingFunctionConfig):
            try:
                return embedding_functions.OpenAIEmbeddingFunction(api_key=config.api_key, model_name=config.model_name)
            except Exception as e:
                raise ImportError(
                    f"Failed to create OpenAI embedding function with model '{config.model_name}'. "
                    f"Ensure openai is installed and API key is valid. Error: {e}"
                ) from e

        elif isinstance(config, CustomEmbeddingFunctionConfig):
            try:
                return config.function(**config.params)
            except Exception as e:
                raise ValueError(f"Failed to create custom embedding function. Error: {e}") from e

        else:
            raise ValueError(f"Unsupported embedding function config type: {type(config)}")

    def _ensure_initialized(self) -> None:
        """Ensure ChromaDB client and collection are initialized."""
        if self._client is None:
            try:
                from chromadb.config import Settings

                settings = Settings(allow_reset=self._config.allow_reset)

                if isinstance(self._config, PersistentChromaDBVectorMemoryConfig):
                    self._client = PersistentClient(
                        path=self._config.persistence_path,
                        settings=settings,
                        tenant=self._config.tenant,
                        database=self._config.database,

View on GitHub (pinned to 027ecf0a37)

Solutions

  1. Read the trailing 'Error: {e}' — it carries the exception your function raised.
  2. Make the params dict keys exactly match the callable's keyword parameter names.
  3. Test the callable standalone: config.function(**config.params) in a REPL before wiring it into the memory.
  4. Fix the root cause inside your custom embedding function (missing file, bad device, etc.).

Example fix

# before
def make_embedder(model_path): ...
config = CustomEmbeddingFunctionConfig(function=make_embedder, params={"path": "model.bin"})
# after
config = CustomEmbeddingFunctionConfig(function=make_embedder, params={"model_path": "model.bin"})
Defensive patterns

Strategy: try-catch

Validate before calling

# smoke-test the factory before wiring it into memory
try:
    fn = config.function(**config.params)
except Exception as e:
    raise RuntimeError(f"custom embedding factory broken: {e}") from e

Type guard

import inspect

def params_match_signature(func, params: dict) -> bool:
    sig = inspect.signature(func)
    try:
        sig.bind(**params)
        return True
    except TypeError:
        return False

Try / catch

try:
    memory = ChromaDBVectorMemory(config=config)
    await memory.update_context(ctx)
except ValueError as e:
    if "custom embedding function" in str(e):
        # inspect e.__cause__ for the factory's own error and fix params/function
        raise
    raise

Prevention

When it happens

Trigger: Configuring CustomEmbeddingFunctionConfig whose function raises on invocation: mismatched params (unexpected keyword), missing required args, or the function itself erroring (e.g. loading a local model that does not exist).

Common situations: Param names in the config not matching the callable signature; callables expecting positional args; custom embedding code failing at construction (missing model files, wrong device).

Related errors

AI-assisted analysis of microsoft/autogen@027ecf0a37 (2026-08-15). Data as JSON: /api/errors/435fa71076db130f. Report an issue: GitHub.