[New Model]: HuggingFaceM4/siglip-so400m-14-980-flash-attn2-navit

[Beniko95J]

The model to consider.

Hi, may I ask has this model (HuggingFaceM4/siglip-so400m-14-980-flash-attn2-navit) beed supported in vllm? The most related implemented model seems to be Idefics2VisionTransformer, but I have some trouble in loading the weights.

The closest model vllm already supports.

Idefics2VisionTransformer

What's your difficulty of supporting the model you want?

No response

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

[DarkLight1337]

It's not supported yet. Community contributions are welcome!

[Flink-ddd]

Hi @DarkLight1337 , Could you assign me the task of accessing the model? recently I just finished reading this siglip model on transformers.

[DarkLight1337]

Sure, thanks for working on this! You should implement this as a pooling model.

[Flink-ddd]

Hi @DarkLight1337 , okay, I will Refer to this format.