fix ImageContentItem to take base64 string as image.data (#909)

# What does this PR do?

- Discussion in
https://github.com/meta-llama/llama-stack/pull/906#discussion_r1936260819

- image.data should accept base64 string as input instead of binary
bytes, change prompt_adapter to account for that.

## Test Plan

```
pytest -v tests/client-sdk/inference/test_inference.py
```

with test in https://github.com/meta-llama/llama-stack/pull/906

## Sources

Please link relevant resources if necessary.


## Before submitting

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

llama_stack/providers/utils/inference/prompt_adapter.py

@@ -135,7 +135,8 @@ async def interleaved_content_convert_to_raw(
                 else:
                     raise ValueError("Unsupported URL type")
             elif image.data:
-                data = image.data
+                # data is a base64 encoded string, decode it to bytes for RawMediaItem
+                data = base64.b64decode(image.data)
             else:
                 raise ValueError("No data or URL provided")
 
@@ -184,8 +185,10 @@ async def localize_image_content(media: ImageContentItem) -> Tuple[bytes, str]:
 
         return content, format
     else:
-        pil_image = PIL_Image.open(io.BytesIO(image.data))
-        return image.data, pil_image.format
+        # data is a base64 encoded string, decode it to bytes first
+        data_bytes = base64.b64decode(image.data)
+        pil_image = PIL_Image.open(io.BytesIO(data_bytes))
+        return data_bytes, pil_image.format
 
 
 async def convert_image_content_to_url(