Comments (4)
from tts-generation-webui.
I found out that RVC config is automatically set device="mps" if mps is available.
So that's why it shows logs as 没有发现支持的N卡, 使用MPS进行推理, then it caused mismatch of device setting.
I can make it work by adding this hard code here.
https://github.com/rsxdalv/tts-generation-webui/blob/main/src/rvc_tab/rvc_tab.py#L61
if device == "cpu": # Workaround for "slow_conv2d_cpu" not implemented for 'Half'
config.is_half = is_half
config.device = "cpu"
But I guess better solution is to enable mps for the device setting in UI, but I'm not sure if it is possible.
from tts-generation-webui.
I can add mps as an option, but I can't test if it will work outright. For now, here's the code to add it to the dropdown
tts-generation-webui/src/rvc_tab/rvc_tab.py
Line 217 in e14c4ca
python
["cuda:0", "cpu", "mps"], label="Device", value="cuda:0"
python
with gr.Group():
gr.Markdown("### Hubert")
with gr.Row():
device = gr.Dropdown(
["cuda:0", "cpu", "mps"], label="Device", value="cuda:0"
)
is_half = gr.Checkbox(
label="Use half precision model (Depends on GPU support)",
value=False,
)
gr.Button(
value="Clear Hubert (to reload on next generation)",
variant="secondary",
).click(
fn=lambda: inject_hubert(None)
) # type: ignore
from tts-generation-webui.
Thanks! I tried to add the mps option in RVC section, and I got this error
NotImplementedError: The operator 'aten::_weight_norm_interface' is not currently implemented for the MPS device. If you want this op to be added in priority during the prototype phase of this feature, please comment on https://github.com/pytorch/pytorch/issues/77764. As a temporary fix, you can set the environment variable `PYTORCH_ENABLE_MPS_FALLBACK=1` to use the CPU as a fallback for this op. WARNING: this will be slower than running natively on MPS.
It seems weight_norm_interface is not implemented yet in Pytorch, so looks like it is better to wait for the update.
I checked the thread here and found many votes for the implementation, so I hope it will be included soon.
from tts-generation-webui.
Related Issues (20)
- Bark voice clone multiple audio inputs? HOT 1
- Errors when generating-- only with Tortoise HOT 6
- Keeping models in vram all the time HOT 9
- start_linux.sh fails at startup on first install HOT 6
- When trying a fresh install, I get these errors HOT 3
- Check if miniconda is already installed. HOT 5
- Colab installation fails: "Failed to install node_modules" HOT 8
- Can rmvpe be added to pitch collection method in RVC? HOT 4
- Bark Voice Generation error HOT 11
- WebUI does not render properly - compared to images I've seen. And tortoise fails to generate HOT 2
- Not able to generate - tried Bark, Tortoise, and MusicGen HOT 17
- Torch version mixup? HOT 6
- Where is gradio ui HOT 3
- IndexError: index 4 is out of range error when checked Multi-band Diffusion HOT 5
- Feature request : Add batch generation to MusicGen HOT 6
- SeamlessM4T Using audio files to implement translation !Can you support it? HOT 3
- Port Connection Issues After Update HOT 2
- Regarding the problem of SeamlessM4T translating and cloning timbre/spoken language, here are cases HOT 5
- Dead "NEW React UI" link HOT 6
- delete this HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from tts-generation-webui.