w-okada / voice-changer

リアルタイムボイスチェンジャー Realtime Voice Changer
Other
15.97k stars 1.73k forks source link

Not using any GPU, only using CPU and Memory. #791

Closed Extra-Time closed 1 year ago

Extra-Time commented 1 year ago

Issue Type

Bug Report

vc client version number

MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.13.zip

OS

Windows 10

GPU

NVIDIA GFORCE RTX 3080

Clear setting

yes

Sample model

yes

Input chunk num

yes

Wait for a while

The GUI successfully launched.

read tutorial

yes

Extract files to a new folder.

yes

Voice Changer type

RVC

Model type

pyTorch

Situation

The w-okoda real time voice changer client kept using CPU, although I set it up as a GPU.

I tried updating, I tried reinstalling, I tried using a different program to help this, I reinstalled the CUDA drivers, I tried everything...

image

it still uses CPU.. help

application window capture

image

logs on terminal

C:\VOICE CHANGER\MMVCServerSIO>MMVCServerSIO.exe -p 18888 --https false --content_vec_500 pretrain/checkpoint_best_legacy_500.pt --content_vec_500_onnx pretrain/content_vec_500.onnx --content_vec_500_onnx_on true --hubert_base pretrain/hubert_base.pt --hubert_base_jp pretrain/rinna_hubert_base_jp.pt --hubert_soft pretrain/hubert/hubert-soft-0d54a1f4.pt --nsf_hifigan pretrain/nsf_hifigan/model --crepe_onnx_full pretrain/crepe_onnx_full.onnx --crepe_onnx_tiny pretrain/crepe_onnx_tiny.onnx --rmvpe pretrain/rmvpe.pt --model_dir model_dir --samples samples.json Booting PHASE :main PYTHON:3.10.11 (tags/v3.10.11:7d4cc5a, Apr 5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)] Activating the Voice Changer. [Voice Changer] download sample catalog. samples_0003_t2.json [Voice Changer] download sample catalog. samples_0003_o2.json [Voice Changer] download sample catalog. samples_0003_d2.json [Voice Changer] model_dir is already exists. skip download samples. Internal_Port:18888 protocol: HTTP


Please open the following URL in your browser.
http://<IP>:<PORT>/
In many cases, it will launch when you access any of the following URLs.
http://127.0.0.1:18888/

[VCClient] Access http://127.0.0.1:18888/ [VCClient] wait web server...0 http://127.0.0.1:18888/ Booting PHASE :main Booting PHASE :MMVCServerSIO [Voice Changer] VoiceChangerManager initializing...

Extra-Time commented 1 year ago

60.4% on CPU and 0.3% on GPU is crazy

w-okada commented 1 year ago

set index to 0.

Extra-Time commented 1 year ago

set index to 0.

I tried, except it's just using less CPU, and still is not using any GPU.

image-62 image-97

This is before and after of index. I just don't think using 0.1% ~ 0.5% GPU is possible though??

I honestly don't know.

Thanks for the reply though!

Extra-Time commented 1 year ago

+my 'wallpaper engine' is using more GPU than my voice changer 😔😔

w-okada commented 1 year ago

set index 0

Extra-Time commented 1 year ago

set index 0

was at 0 already, check this screenshot out image

ralyx commented 1 year ago

I have amd graphics card but i had the same issue i solved it by exporting the custom voice to onnx ,uploading it to the slot and then switching to another character and then to that character again.And if it doesn't work you try every extra and chunk combination.I don't actually have a lot of knowledge but it's worth trying.And i also changed a setting in advanced settings.I changed a value from 100 to 300 hope it works for you ;).

Extra-Time commented 1 year ago

I tried voice onnx feature, and it didnt work.. also advanced setting one was at 300 at all costs :) thanks tho

w-okada commented 1 year ago

Does the same phenomenon occur with the sample model too?

w-okada commented 1 year ago

no res close