DAMO-NLP-SG / DAMO-NLP-SG/VideoLLaMA2
AVQA Result Different from The Reported Result.
Open
- Dominant language
- Python
- Stars
- 1.3k
- Forks
- 90
- PR merge metrics
- No merged PRs in 30d
Description
Hi,
Thank you for the great work.
I'm using `audio_visual` branch with [VideoLLaMA2.1-7B-AV](https://huggingface.co/DAMO-NLP-SG/VideoLLaMA2.1-7B-AV), Audio Encoder: [Fine-tuned BEATs_iter3+(AS2M)(cpt2](https://1drv.ms/u/s!AqeByhGUtINrgcpj8ujXH1YUtxooEg?e=E9Ncea), video encoder: [siglip-so400m-patch14-384](https://huggingface.co/google/siglip-so400m-patch14-384).
The result found: `Overall Accuracy: 65.75% Total Responses: 9185`
Which is significantly lower than the reported result **80.9**.
Please give me some suggestions on how to reproduce the results. Or what I'm doing incorrectly.
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.