-
float32/amp dynamo benchmark runs on triton master
cc @ezyang @soumith @msaroufim @wconstab @ngimel @bdhirsh
-
**Describe**
Model I am using (UniLM, MiniLM, LayoutLM ...): BEiT
Congrats on your awesome work! I have two questions regarding reproducing the segmentation result on ADE20k.
1. The paper menti…
-
-
Hi, I would like to port the "MMsegmentation" video-demo.py to ViT-Adapter and run it, but it stops.
I already have
`cd segmentation/
ln -s . /detection/ops . /
cd ops & sh make.sh`
has been do…
-
Hi,
May I ask when will the code and pretrainde model of BEIT-3 be released?
Thanks
-
Hello,
Thank you for your great work BEIT. I have a question about the loss. In the section 2.3 of BEIT paper, you say the pre-training objective is definded in formula 1. In formula 1, it seems the …
-
**Describe**
BEiT3
-
I am very interested in the BEIT v3 paper. When will the code be publicly available? Thanks!
-
Thanks for your inspiring and effective work. You achieved really great performance on zero-shot and finetune retrieval. But we are wondering how the batch size could be 16k in intermediate finetune…
-
**Describe**
The model I am asking BEIT3
Dear authors,
Thank you for such seminal work!
I am wondering about the pretraining time and the required hardware resource. If I could be wrong, I d…