bytedance / bytedance/MegaTTS3
windows下运行webui成功了，但是加载cuda会报错，只能用cpu
Open
- Dominant language
- Python
- Stars
- 6.1k
- Forks
- 474
- PR merge metrics
- No merged PRs in 30d
Description
conda下pynini==2.1.6，WeTextProcessing==1.0.4.1，用conda install -c pytorch pytorch torchvision安装的最新版。torch.cuda.amp.autocast(dtype=self.precision, enabled=True)替换成了torch.amp.autocast(device_type='cuda',dtype=self.precision, enabled=True)。但运行时只能cpu推理，无法加载cuda。
Contributor guide
No contributing guide indexed for this repository
Research direction
Start by reproducing the Windows conda setup with pynini==2.1.6, WeTextProcessing==1.0.4.1, and the latest PyTorch installed from the pytorch channel. Inspect the CUDA loading and autocast behavior shown in the issue; the work is done when WebUI inference uses CUDA successfully instead of falling back to CPU.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100