Hybrid Chunker utilizing IBM watsonx.ai embedding model installed

[helloooojoy]

hi there, is there example of use IBM watsonx.ai client/function for docling hybrid chunker tokenizer? I understand below code snippet will work.

MAX_TOKENS =5000
EMBED_MODEL_ID="ibm-granite/granite-embedding-125m-english"
tokenizer = HuggingFaceTokenizer(
        tokenizer=AutoTokenizer.from_pretrained(EMBED_MODEL_ID),
        max_tokens=MAX_TOKENS,  # optional, by default derived from `tokenizer` for HF case
    )

However, I'm utilizing watsonx.ai on-prem instance and would like to use our installed embedding model instead of downloading from huggingface. Is there example code snippet about how this will work? I tried with embedding function and it errors out as below.

TypeError: Can't instantiate abstract class BaseChunker with abstract methods from_dict, split_documents, to_dict
Any help is appreciated! Thanks!