Feng Li

Results 119 comments of Feng Li

You should put the tensor on the same device, GPU or CPU.

Sorry for the late reply. We have not released the code yet. The implementation is easy, and you can add noise only to the 2D anchors.

Yes, the training data is under preparation. We will release it soon.

Can you give a link for the work Cambricon?

Yes, we have just released the data.

Yes, we have released the 0.5b model [here](https://huggingface.co/lmms-lab/llava-next-interleave-0.5b)

If you want a interleaved demo, you can try our benchmark inference code.

Thanks for pointing out this problem. The \ prompt should be used in the value of gpt. We will fix it recently.