Files
exo/.mlx_typings/mlx_vlm/prompt_utils.pyi
T
rltakashige 635801d515 Add multimodality! (#1802)
## Motivation

Images!

TODO (in a future PR): Add audio and video support.

## Test Plan

### Manual Testing
<img width="2652" height="1900" alt="image"
src="https://github.com/user-attachments/assets/7d3a7137-542f-4f94-9193-2c73b7c4a5ec"
/>

<img width="2770" height="1956" alt="image"
src="https://github.com/user-attachments/assets/e3c3a096-8029-4409-97a6-aca31a9a3f24"
/>
<img width="2738" height="1768" alt="image"
src="https://github.com/user-attachments/assets/d70ea37f-cd1d-4a4c-ad08-3beb9fafa380"
/>

(And batching also works)

---------

Co-authored-by: David Hind <davehind@yahoo.co.uk>
2026-03-30 11:52:19 +01:00

13 lines
274 B
Python

from typing import Any
def get_message_json(
model_name: str,
prompt: str,
role: str = "user",
skip_image_token: bool = False,
skip_audio_token: bool = False,
num_images: int = 0,
num_audios: int = 0,
**kwargs: Any,
) -> dict[str, Any]: ...