Zhe Chen
Hi, you need to replace `nn.SyncBatchNorm` with `nn.BatchNorm2d` in `adapter_modules.py` when training the model using 1 GPU.
> Hi, I try to work out a usable EMA module with Zero Stage 3. See below: > > ```python > from deepspeed.runtime.zero import GatheredParameters > > class DSEma(nn.Module): >...
Hi, you can set device_map='auto' to use multiple GPUs for inference. May I ask if you are currently meeting out-of-memory issues with 8 V100 GPUs without Flash Attention? ``` path...
Thanks for your feedback. Could you please try to see if the model can be used normally with only 4 GPUs?
Sorry for the late reply. I would like to ask if you installed Flash Attention on the v100 machine. If so, could you please uninstall Flash Attention and try again?...
> > Sorry for the late reply. I would like to ask if you installed Flash Attention on the v100 machine. > > If so, could you please uninstall Flash...
> Update: I changed the version of mmvc-full and followed the README.md provided and everything works well! > > The link to the notebook is in the previous comment, Ill...
> It's okay. Thanks for reply. Hi, [Multimodal Transfer](https://github.com/FeliMe/multimodal_style_transfer) is ready now. https://github.com/czczup/URST/tree/main/Wang2017Multimodal
FAST的模型以及代码将在近期开源，链接如下～ The models and code of FAST will be released in https://github.com/czczup/FAST
FAST的模型以及代码将在近期开源，链接如下～ https://github.com/czczup/FAST