Chaoyou Fu comments

Results 14 comments of


                                            Chaoyou Fu

Can't run this demo on hugging face

The problem has been fixed. Please try again. Thank you!

Gpu required?

2 48G GPUs

关于audio模态输出问题

Yes

How many samples are in the dataset?

Hello, the socre of 2800 (perception 2000 + cognition 800) is only related to the number of subtasks (10 + 4, each task has a score of 200, including acc...

how to set eval dataset

Hi, [VLMEvalKit](https://github.com/open-compass/VLMEvalKit) of OpenCompass has supported our both VITA-1.5 and VITA-1.0 models.

Release VITA-1.5 on Hugging Face

Hi Niels! I'm sorry I just got your message. I am very happy to deploy VITA-1.5 on Huggingface Spaces. But our model needs one GPU with 80GB memory. Is this...

各位大佬，这个要多少显存才能跑呀，interactive demo的，可以指定多张卡吗

2 48G GPUs or 1 80G GPU

Is the mobile demo running fully on-device or using cloud inference?

Hi, the model is running on cloud GPUs.

Bad performances on English question in interactive demo.

Hi! Thanks for your attention on our work! You could try our new VITA-1.5 version, which has much stronger speech performance.

use gradio webrtc for real time app

Hi, thanks for your suggestion!