Xuan-Phi Nguyen comments

Results 25 comments of


                                            Xuan-Phi Nguyen

Better llava next.

@messlav Can you try turning this line back to `new_token_positions += new_token_positions[:, -1].max() - new_token_positions[:, -1:]` ? https://github.com/huggingface/transformers/blob/a0102a425dc8d01fddf215444aa2e54dfd8b7eb2/src/transformers/models/llava_next/modeling_llava_next.py#L565

The image uploaded in the multimodal chatbot might be too small to be clearly visible.

@yvrjsharma You can pass in the root gr.Blocks with `css=` e.g: ``` with gr.Blocks(css=".message-wrap.svelte-1lcyrx4>div.svelte-1lcyrx4 img {min-width: 200px}") as demo: ``` Note: it only works for the root block.

Multimodal Textbox (Chat Input Component)

@dawoodkhan82 Many thanks for this great features. However, how to use this with `gr.Examples` ? I need the specify and `gr.Examples` so that when clicking on an example, it will...

CUDA 12.1 vllm==0.2.3 Double Free

When sampling multiple times, the results tend to become significantly longer.

+1 It looks like the model fails to generate eos in the subsequent generations. `n` here seems to be applied for beam search, where 2 sample generation cannot be the...