Shikun Liu

Typing: \begin{align_} [I usually typed the brackets first, then the environment name, in multiselection mode] \begin{align_} Autocomplete[should've] : \begin{align_} \end{align_} \begin{align_} \end{align_} After Autocomplete[in reality]: \begin{align* \end{align_}} \begin{align_ \end{align*}} _Problem...

> Hello, thanks for your nice work! I am now having trouble reproducing the reported score on the VQA task. I evaluated the checkpoint downloaded from https://storage.googleapis.com/sfr-vision-language-research/BLIP/models/model_base_vqa_capfilt_large.pth and followed the...

@sdc17 Thanks for the update. At least we have confirmed that we reproduced a consistent performance.

I would suggest checking out: https://github.com/huggingface/accelerate which should be relatively easy to deploy any model in a distributed setting.

Well...Maybe move utils/__init__.py to utils.py? See whether this will work? Otherwise you may install the entire package from https://github.com/pmeletis/panoptic_parts. You can clearly see the function defined in init.. so just...

Sorry I am not able to reproduce your problem. It seems that your data files are wrong. Could you just pip install panoptic_parts, and make sure the CityScapes files are...

Thanks, that is very helpful. Here are some follow-up questions. 1. Well...That's interesting. But you can load the weights, and modify the key name, and then save the weights. You...

Hello again, Thanks for your clarification. For Q4, I think that I have found an elegant solution to this: For any random index, (to reshuffle the style in the image-level),...

Update: We need to swap the index for fake_pred as well. Further, for any non-defined style code swapped to a segmentation map with the missing region, it would cause instability...

Check out this work: https://arxiv.org/abs/2304.08818