Whether to truncate the input texts to fit within the context length.
If True, over-length input texts will be truncated to fit within the context length, before vectorized by the embedding model. If False, an error will be raised if any given text exceeds the context length.
truncation: bool = True