add image support to NVIDIA inference provider (#907)

# What does this PR do?

add support to the NVIDIA Inference provider for image inputs


## Test Plan

1. Run local [Llama 3.2 11b vision
instruct](https://build.nvidia.com/meta/llama-3.2-11b-vision-instruct?snippet_tab=Docker)
NIM
2. Start a stack, e.g. `llama stack run
llama_stack/templates/nvidia/run.yaml --env
NVIDIA_BASE_URL=http://localhost:8000`
3. Run image tests, e.g. `LLAMA_STACK_BASE_URL=http://localhost:8321
pytest -v tests/client-sdk/inference/test_inference.py
--vision-inference-model meta-llama/Llama-3.2-11B-Vision-Instruct -k
image`


## Before submitting

- [ ] This PR fixes a typo or improves the docs (you can dismiss the
other checks if that's the case).
- [x] Ran pre-commit to handle lint / formatting issues.
- [x] Read the [contributor
guideline](https://github.com/meta-llama/llama-stack/blob/main/CONTRIBUTING.md),
      Pull Request section?
- [ ] Updated relevant documentation.
- [x] Wrote necessary unit or integration tests.

llama_stack/providers/utils/inference/prompt_adapter.py

@@ -186,6 +186,7 @@ async def localize_image_content(media: ImageContentItem) -> Tuple[bytes, str]:
         return content, format
     else:
         # data is a base64 encoded string, decode it to bytes first
+        # TODO(mf): do this more efficiently, decode less
         data_bytes = base64.b64decode(image.data)
         pil_image = PIL_Image.open(io.BytesIO(data_bytes))
         return data_bytes, pil_image.format