[BUG] Indexing multimodal with images using text-only models raises 500

[vicilliar]

Describe the bug
When indexing multimodal data with images into a text model (hf/all_datasets_v4_MiniLM-L6) raises a 500 internal error.

Marqo output:

    image_vectors = s2_inference.vectorise(
  File "/app/src/marqo/s2_inference/s2_inference.py", line 74, in vectorise
    vector_batches.append(_convert_tensor_to_numpy(available_models[model_cache_key][AvailableModelsKey.model].encode(batch, normalize=normalize_embeddings, **kwargs)))
  File "/app/src/marqo/s2_inference/hf_utils.py", line 138, in encode
    inputs = self.tokenizer(sentence, padding=True, truncation=True, max_length=self.max_seq_length,
  File "/usr/local/lib/python3.8/dist-packages/transformers/tokenization_utils_base.py", line 2548, in __call__
    encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs)
  File "/usr/local/lib/python3.8/dist-packages/transformers/tokenization_utils_base.py", line 2606, in _call_one
    raise ValueError(
ValueError: text input must of type `str` (single example), `List[str]` (batch or single pretokenized example) or `List[List[str]]` (batch of pretokenized examples).

To Reproduce
Steps to reproduce the behavior:

1. Create a marqo index with model hf/all_datasets_v4_MiniLM-L6
2. Index documents in multimodal form with images
3. Observe error in marqo and client

Expected behavior
Marqo should return a 400 with a better message explaining why the error happened.

Screenshots
Client error:

Marqo error:

Desktop (please complete the following information):

• SageMaker notebook client