vik
Thanks! The model currently only sees a 378x378 version of any images fed in, so it's not going to be able to see any fine grained details. I don't want...
It is available now! You may need to use the latest pre-release version of ollama. https://ollama.com/library/moondream
Hey - not planning to release details about the dataset at this time. If your company needs this information to be able to deploy it please reach out to [email protected]...
Can you check if you're using the `2024-03-06` revision? Previous ones don't support it.
`answer_question` only supports a single image, there's a new `batch_answer` API for batch generation: ``` answers = moondream.batch_answer( images=[Image.open(''), Image.open('')], prompts=["Describe this image.", "Are there people in this image?"], tokenizer=tokenizer,...
Is `enc_images` a list of PIL Images? It looks like you might have already run them through the image encoder?
Nice! This is good feedback, I'm going to improve the error messages.
Looks like you have an old version of timm. Try `pip install -U timm`
Going to close this and use #65 as the tracking issue.
Related issue: https://github.com/vikhyat/moondream/issues/64 This is due to using an old version of `timm` from before when it added support for SigLIP. You can view the version of timm in your...