Video-LLaMA

Video-LLaMA copied to clipboard

nvrtc: error: invalid value for --gpu-architecture (-arch)

BinZhu-ece

Repeating itself

2

Hi, Thanks for the code! For multiple videos we found that the model is quite often just repeating itself when answering. For instance, it might make descriptions like "In this...

ThierryDeruyttere

Hi, thanks for the great work! While reading the code, I noticed that you have used self-implemented version of BLIP and BERT etc. as oppose to directly importing the corresponding...

MoonBlvd

Thanks for your contributions to VideoLLaMA! It is an impressive work. I have just fixed one bug about resuming checkpoint. In `video_llama/runners/runner_base.py` , The position of parameter `strict=False` seems to...

CyrilSterling

Thanks for your contributions to VideoLLaMA! It is an impressive work. I have just fixed some details below. 1. `video_llama/runners/runner_base.py` has a bug for resuming the model. The position of...

CyrilSterling

How much time did you take to complete this program. What was the most difficult part that you came across?

Zahidpichen1

https://github.com/DAMO-NLP-SG/Video-LLaMA/blob/main/train_configs/visionbranch_stage1_pretrain.yaml#L73 这一行设置一个epoch跑2500个iters，数据集下载完是2253277个，batchsize 32 8张卡应该是8801个iters，是不是不太对呀

xmy0916

您好！ 大佬能给一个access model 生成的answer的logit的脚本或者方法吗？

Anthony6197

visionbranch_stage1预训练webvid数据集问题

1

我在使用webvid-2M数据集预训练时，出现了如下报错： 2023-09-06 10:50:10,790 [INFO] number of trainable parameters: 19544832 2023-09-06 10:50:10,793 [INFO] Start training epoch 0, 2500 iters per inner epoch. [10:50:14] /home/spillai/miniconda3/envs/decord/conda-bld/decord_1631735293613/work/src/video/video_reader.cc:83: ERROR opening: /data3/ryb/webvid/data/videos/000051_000100/1066651354.mp4, No such file or...

RANYABING

I hit one issue when running inference.py with Billa7B-SFT. It runs well on languague model vicuna7B while I hit this erorr when the languague model gets changed to Billa7B-SFT. Here...

dongxiao92