A卡so-vits-svc3.0煉丹教程
前言：
本教程僅適用于A卡，使用N卡的您應(yīng)當(dāng)前往soVITS3.0煉丹教程。本人并不是專業(yè)人員，僅業(yè)余愛好者，有錯誤歡迎指正。本人參考過的所有文章鏈接都放在文章末尾。您應(yīng)當(dāng)知道煉丹對顯卡的傷害不小于挖礦！
前期準(zhǔn)備：
Ubuntu22.04、Python3.7.0
推薦使用Anaconda管理虛擬環(huán)境。系統(tǒng)盤不應(yīng)小于50G。您不應(yīng)該使用虛擬機煉丹。
#打上顯卡驅(qū)動
wget https://repo.radeon.com/amdgpu-install/22.40/ubuntu/jammy/amdgpu-install_5.4.50401-1_all.deb
sudo dpkg -i amdgpu-install_5.4.50401-1_all.deb
sudo apt update
sudo apt upgrade -y
sudo amdgpu-install --no-dkms
sudo apt install rocm-dev
sudo apt install radeontop（此項非必選）
sudo apt-get install libbz2-dev
sudo reboot
#配置環(huán)境
ls -l /dev/dri/render*
sudo usermod -a -G render $LOGNAME
sudo usermod -a -G video $LOGNAME
sudo reboot
#測試環(huán)境
rocm-smi
/opt/rocm/bin/rocminfo
/opt/rocm/opencl/bin/clinfo
沒報錯即可
PIP安裝依賴（事故多發(fā)地）
#所有指令均應(yīng)當(dāng)在項目文件夾的終端下執(zhí)行。
#請先執(zhí)行
python -m pip install --upgrade pip
pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
sudo apt-get install libasound-dev portaudio19-dev libportaudio2 libportaudiocpp0
sudo apt-get install libbz2-dev
#防止后面安裝pyaudio報錯，同時加快pip的安裝速度
（不推薦）你對自己處理依賴庫有信心可以直接復(fù)制下面的指令安裝。
pip install -r requirements.txt -i https://pypi.mirrors.ustc.edu.cn/simple/
（推薦)自己動手一個一個安裝
pip install librosa==0.8.1
pip install Flask==2.1.2
pip install Flask_Cors==3.0.10
pip install gradio==3.4.1
pip install
pip install
pip install
pip install
pip install
pip install
pip install
pip install
pip install
pip install
pip install
pip install
pip install praat-parselmouth
pip install tensorboard
pip install onnx
pip install onnxsim
pip install onnxoptimizer
安裝MIopen
sudo apt-get install miopen-hip
下載https:www.123pan.com/xW39-oyMmH
sudo dpkg -i miopenkernels-gfx1030-36kdb_1.1.0.50200-65_amd64.deb
安裝pytorch
RDNA2架構(gòu)的顯卡：pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/rocm5.2
RX580：pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/rocm3.7
到這里前期工作基本完成！
部署項目
所有指令均應(yīng)當(dāng)在項目文件夾的終端下執(zhí)行。
#安裝預(yù)訓(xùn)練模型
wget -P hubert/ https://github.com/bshall/hubert/releases/download/v0.1/hubert-soft-0d54a1f4.pt
wget -P logs/32k/ https://huggingface.co/innnky/sovits_pretrained/resolve/main/G_0.pth
wget -P logs/32k/ https://huggingface.co/innnky/sovits_pretrained/resolve/main/D_0.pth
#數(shù)據(jù)預(yù)處理
重采樣至 32khz
python resample.py
自動劃分訓(xùn)練集 驗證集 測試集 以及自動生成配置文件
HSA_OVERRIDE_GFX_VERSION=10.3.0 python preprocess_flist_config.py
生成hubert與f0
HSA_OVERRIDE_GFX_VERSION=10.3.0 python preprocess_hubert_f0.py
#訓(xùn)練
HSA_OVERRIDE_GFX_VERSION=10.3.0 python train.py -c configs/config.json -m 32k
#推理
HSA_OVERRIDE_GFX_VERSION=10.3.0 python inference_main.py
#訓(xùn)練監(jiān)控
tensorboard --logdir=logs/32k
瀏覽器打開http://127.0.0.1:6006即可查看訓(xùn)練情況
rocm-smi
終端輸入即可查看顯卡運行情況
本文參考的連接：
https://blog.csdn.net/qq_44948500/article/details/127346390?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522167708346516800213013835%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fall.%2522%257D&request_id=167708346516800213013835&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~all~first_rank_ecpm_v1~rank_v31_ecpm-3-127346390-null-null.142^v73^wechat_v2,201^v4^add_ask,239^v2^insert_chatgpt&utm_term=xiaozhou&spm=1018.2226.3001.4187
https://www.bilibili.com/read/cv20500632?spm_id_from=333.999.0.0
https://www.bilibili.com/read/cv20997087?spm_id_from=333.999.0.0
https://github.com/innnky/so-vits-svc
https://www.bilibili.com/read/cv20533940?spm_id_from=333.999.0.0
https://www.bilibili.com/video/BV1fD4y1w7KX/?spm_id_from=333.999.0.0&;vd_source=305e43fe4b1fa2a34139fec29c9becdf
so-vits-svc4.0的部署方法基本相同，制作不易，求個贊～
