wangxinliang
wangxinliang
Hello, can you share vcoco.pickle in generate_vcoco_official.py? Thanks.
Thank you for the excellent work. Can you provide the code of swin-t-adapter?
Hi, thx for your great work. Can you share the method to draw the figure 2 in your PVT paper?
Can you provide the 'pretrained/dinov2/dinov2_vitl14_pretrain'?
The finetune data for InternVL-Chat-v1.2 used 1.2M open-source data. Could you please specify what the 12M finetune data for v1.2 plus consists of?
Would it be possible to enhance the detection capability of InternVL by incorporating more data combined with grounding instructions during the fine-tuning stage?
Why does the file internvl_chat_v1_2_hermes2_yi34b_448_finetune.sh include --freeze_backbone False? Isn't the visual encoder supposed to be frozen during the pre-training phase?
The demo raised an error: 'scaled_dot_product_attention() got an unexpected keyword argument 'scale'. Can you fix it?
Can you share the finetune.sh and pretrain.sh to train TinyLLaVA-1.5B?
python: 3.8 torch: pip install torch==1.9.1+cu111 torchvision==0.10.1+cu111 torchaudio==0.9.1 -f https://download.pytorch.org/whl/torch_stable.html mmcv: https://download.openmmlab.com/mmcv/dist/cu111/torch1.9.0/index.html [./mmcv_full-1.7.2-cp38-cp38-manylinux1_x86_64.whl](https://download.openmmlab.com/mmcv/dist/cu111/torch1.9.0/mmcv_full-1.7.2-cp38-cp38-manylinux1_x86_64.whl)