chattts一步步的记录，先跑起来。
0.下载git工具
Git - Downloads (git-scm.com)
https://git-scm.com/downloads
Download – TortoiseGit – Windows Shell Interface to Githttps://tortoisegit.org/download/
1.安装 随意，可以安汉化，也可不安。无所谓
2.建个目录，我的上I:chat_kimi,你随意
3.打开官方
https://github.com/2noise/ChatTTS拉取链接
https://github.com/2noise/ChatTTS.git
4.假设你已经安装了conda.
conda create --name chat_kimi python=3.11conda activate chat_kimi
如下即可：
或者通过界面方式创建也可。
5安装所要的文件，仅参考
(base) PS C:\Users\dell> conda activate chat_kimi
(chat_kimi) PS C:\Users\dell> cd I:\chat_kimi
(chat_kimi) PS I:\chat_kimi> cd chattts
用国内的源：
pip install -r requirements.txt -i https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple 
6.模型Hf国内源，有一个位置是修改这个自动下载地址，等我找到后再更新到这。
HF-Mirrorhttps://hf-mirror.com/结构如
新建一个文件夹 models.或者，直接放在 chattts目录下，如：第二图：第二图的好处是所有代码不用修改了，直接用。
将网站上的文件全部下载到对应的位置
下载后，注意个别文件的主文件名，给加上了路径名了，一定要改回去。
对应下载到相应的地方。大约20分钟左右。
7.修改模型路径位置。如果是第一图的话.
注意这个位置。
7.运行一下。
(chat_kimi) PS I:\chat_kimi\ChatTTS> python examples\web\web.py
C:\Users\dell\.conda\envs\chat_kimi\python.exe: can't open file 'I:\\chat_kimi\\ChatTTS\\examples\\web\\web.py': [Errno 2] No such file or directory
(chat_kimi) PS I:\chat_kimi\ChatTTS> python examples\web\webui.py
[+0800 20241013 17:03:24] [WARN] WebUI | funcs | no ffmpeg installed, use wav file output
[+0800 20241013 17:03:24] [INFO] WebUI | webui | loading ChatTTS model...
[+0800 20241013 17:03:24] [INFO] ChatTTS | dl | checking assets...
[+0800 20241013 17:03:25] [INFO] ChatTTS | dl | all assets are already latest.
[+0800 20241013 17:03:25] [WARN] ChatTTS | gpu | no GPU found, use CPU instead
[+0800 20241013 17:03:25] [INFO] ChatTTS | core | use device cpu
[+0800 20241013 17:03:25] [INFO] ChatTTS | core | vocos loaded.
[+0800 20241013 17:03:25] [INFO] ChatTTS | core | dvae loaded.
[+0800 20241013 17:03:26] [INFO] ChatTTS | core | embed loaded.
[+0800 20241013 17:03:26] [INFO] ChatTTS | core | gpt loaded.
[+0800 20241013 17:03:26] [INFO] ChatTTS | core | speaker loaded.
[+0800 20241013 17:03:26] [INFO] ChatTTS | core | decoder loaded.
[+0800 20241013 17:03:26] [INFO] ChatTTS | core | tokenizer loaded.
[+0800 20241013 17:03:26] [WARN] WebUI | funcs | Package nemo_text_processing not found!
[+0800 20241013 17:03:26] [WARN] WebUI | funcs | Run: conda install -c conda-forge pynini=2.1.5 && pip install nemo_text_processing
[+0800 20241013 17:03:26] [WARN] WebUI | funcs | Package WeTextProcessing not found!
[+0800 20241013 17:03:26] [WARN] WebUI | funcs | Run: conda install -c conda-forge pynini=2.1.5 && pip install WeTextProcessing
[+0800 20241013 17:03:26] [INFO] WebUI | webui | Models loaded successfully.
* Running on local URL: http://0.0.0.0:8080To create a public link, set `share=True` in `launch()`.
text: 0%|▏ | 1/384(max) [00:00, 3.95it/s]We detected that you are passing `past_key_values` as a tuple of tuples. This is deprecated and will be removed in v4.47. Please convert your cache or use an appropriate `Cache` class (https://huggingface.co/docs/transformers/kv_cache#legacy-cache-format)
text: 17%|█████████████▎ | 67/384(max) [00:05, 13.30it/s]
code: 24%|█████████████████▉ | 495/2048(max) [00:31, 15.72it/s]8.gpu竟然不干活。发生了什么？
CUDA Toolkit Archive | NVIDIA Developerhttps://developer.nvidia.com/cuda-toolkit-archive
CUDA安装教程（超详细）-CSDN博客https://blog.csdn.net/m0_45447650/article/details/123704930?ops_request_misc=%257B%2522request%255Fid%2522%253A%252229DBE75F-820D-4C34-94EB-A83EC55EC789%2522%252C%2522scm%2522%253A%252220140713.130102334..%2522%257D&request_id=29DBE75F-820D-4C34-94EB-A83EC55EC789&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~all~top_positive~default-1-123704930-null-null.142%5Ev100%5Epc_search_result_base5&utm_term=cuda%E5%AE%89%E8%A3%85&spm=1018.2226.3001.4187
cuda安装
安装cuda时，第一次会让设置临时解压目录，第二次会让设置安装目录；临时解压路径，建议默认即可，也可以自定义。安装结束后，临时解压文件夹会自动删除；安装目录，建议默认即可；注意：临时解压目录千万不要和cuda的安装路径设置成一样的，否则安装结束，会找不到安装目录的！！！选择自定义安装安装完成后，配置cuda的环境变量；命令行中，测试是否安装成功；双击“exe文件”，选择下载路径（推荐默认路径）验证：
cuDNN Archive | NVIDIA Developerhttps://developer.nvidia.com/rdp/cudnn-archive
最详细！Windows下的CUDA与cuDNN详细安装教程_windows安装cuda和cudnn-CSDN博客https://blog.csdn.net/weixin_52677672/article/details/135853106?ops_request_misc=%257B%2522request%255Fid%2522%253A%252229DBE75F-820D-4C34-94EB-A83EC55EC789%2522%252C%2522scm%2522%253A%252220140713.130102334..%2522%257D&request_id=29DBE75F-820D-4C34-94EB-A83EC55EC789&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~all~top_positive~default-2-135853106-null-null.142%5Ev100%5Epc_search_result_base5&utm_term=cuda%E5%AE%89%E8%A3%85&spm=1018.2226.3001.4187
Win10安装ChatTTS-2024-cuda10.1_window 10 chattts安装-CSDN博客https://blog.csdn.net/counsellor/article/details/141437597?ops_request_misc=&request_id=&biz_id=102&utm_term=chattts%20pip%20cuda&utm_medium=distribute.pc_search_result.none-task-blog-2~all~sobaiduweb~default-2-141437597.142%5Ev100%5Epc_search_result_base5&spm=1018.2226.3001.4187
conda install pytorch==2.4.0 torchvision==0.19.0 torchaudio==2.4.0 pytorch-cuda=12.4 -c pytorch -c nvidia
漫长的等待。注意硬盘的空间。
(chat_kimi) PS I:\chat_kimi\ChatTTS> conda install pytorch==2.4.0 torchvision==0.19.0 torchaudio==2.4.0 pytorch-cuda=12.4 -c pytorch -c nvidia
Channels:- pytorch- nvidia- defaults
Platform: win-64
Collecting package metadata (repodata.json): done
Solving environment: done## Package Plan ##environment location: C:\Users\dell\.conda\envs\chat_kimiadded / updated specs:- pytorch-cuda=12.4- pytorch==2.4.0- torchaudio==2.4.0- torchvision==0.19.0The following packages will be downloaded:package | build---------------------------|-----------------pytorch-2.4.0 |py3.11_cuda12.4_cudnn9_0 1.20 GB pytorch------------------------------------------------------------Total: 1.20 GBThe following NEW packages will be INSTALLED:blas pkgs/main/win-64::blas-1.0-mklbrotli-python pkgs/main/win-64::brotli-python-1.0.9-py311hd77b12b_8certifi pkgs/main/win-64::certifi-2024.8.30-py311haa95532_0charset-normalizer pkgs/main/noarch::charset-normalizer-3.3.2-pyhd3eb1b0_0cuda-cccl nvidia/win-64::cuda-cccl-12.6.77-0cuda-cccl_win-64 nvidia/noarch::cuda-cccl_win-64-12.6.77-0cuda-cudart nvidia/win-64::cuda-cudart-12.4.127-0cuda-cudart-dev nvidia/win-64::cuda-cudart-dev-12.4.127-0cuda-cupti nvidia/win-64::cuda-cupti-12.4.127-0cuda-libraries nvidia/win-64::cuda-libraries-12.4.0-0cuda-libraries-dev nvidia/win-64::cuda-libraries-dev-12.4.0-0cuda-nvrtc nvidia/win-64::cuda-nvrtc-12.4.127-0cuda-nvrtc-dev nvidia/win-64::cuda-nvrtc-dev-12.4.127-0cuda-nvtx nvidia/win-64::cuda-nvtx-12.4.127-0cuda-opencl nvidia/win-64::cuda-opencl-12.6.77-0cuda-opencl-dev nvidia/win-64::cuda-opencl-dev-12.6.77-0cuda-profiler-api nvidia/win-64::cuda-profiler-api-12.6.77-0cuda-runtime nvidia/win-64::cuda-runtime-12.4.0-0cuda-version nvidia/noarch::cuda-version-12.6-3filelock pkgs/main/win-64::filelock-3.13.1-py311haa95532_0freetype pkgs/main/win-64::freetype-2.12.1-ha860e81_0gmpy2 pkgs/main/win-64::gmpy2-2.1.2-py311h7f96b67_0idna pkgs/main/win-64::idna-3.7-py311haa95532_0intel-openmp pkgs/main/win-64::intel-openmp-2023.1.0-h59b6b97_46320jinja2 pkgs/main/win-64::jinja2-3.1.4-py311haa95532_0jpeg pkgs/main/win-64::jpeg-9e-h827c3e9_3lcms2 pkgs/main/win-64::lcms2-2.12-h83e58a3_0lerc pkgs/main/win-64::lerc-3.0-hd77b12b_0libcublas nvidia/win-64::libcublas-12.4.2.65-0libcublas-dev nvidia/win-64::libcublas-dev-12.4.2.65-0libcufft nvidia/win-64::libcufft-11.2.0.44-0libcufft-dev nvidia/win-64::libcufft-dev-11.2.0.44-0libcurand nvidia/win-64::libcurand-10.3.7.77-0libcurand-dev nvidia/win-64::libcurand-dev-10.3.7.77-0libcusolver nvidia/win-64::libcusolver-11.6.0.99-0libcusolver-dev nvidia/win-64::libcusolver-dev-11.6.0.99-0libcusparse nvidia/win-64::libcusparse-12.3.0.142-0libcusparse-dev nvidia/win-64::libcusparse-dev-12.3.0.142-0libdeflate pkgs/main/win-64::libdeflate-1.17-h2bbff1b_1libjpeg-turbo pkgs/main/win-64::libjpeg-turbo-2.0.0-h196d8e1_0libnpp nvidia/win-64::libnpp-12.2.5.2-0libnpp-dev nvidia/win-64::libnpp-dev-12.2.5.2-0libnvfatbin nvidia/win-64::libnvfatbin-12.6.77-0libnvfatbin-dev nvidia/win-64::libnvfatbin-dev-12.6.77-0libnvjitlink nvidia/win-64::libnvjitlink-12.4.99-0libnvjitlink-dev nvidia/win-64::libnvjitlink-dev-12.4.99-0libnvjpeg nvidia/win-64::libnvjpeg-12.3.1.89-0libnvjpeg-dev nvidia/win-64::libnvjpeg-dev-12.3.1.89-0libpng pkgs/main/win-64::libpng-1.6.39-h8cc25b3_0libtiff pkgs/main/win-64::libtiff-4.5.1-hd77b12b_0libuv pkgs/main/win-64::libuv-1.48.0-h827c3e9_0libwebp-base pkgs/main/win-64::libwebp-base-1.3.2-h2bbff1b_0lz4-c pkgs/main/win-64::lz4-c-1.9.4-h2bbff1b_1markupsafe pkgs/main/win-64::markupsafe-2.1.3-py311h2bbff1b_0mkl pkgs/main/win-64::mkl-2023.1.0-h6b88ed4_46358mkl-service pkgs/main/win-64::mkl-service-2.4.0-py311h2bbff1b_1mkl_fft pkgs/main/win-64::mkl_fft-1.3.10-py311h827c3e9_0mkl_random pkgs/main/win-64::mkl_random-1.2.7-py311hea22821_0mpc pkgs/main/win-64::mpc-1.1.0-h7edee0f_1mpfr pkgs/main/win-64::mpfr-4.0.2-h62dcd97_1mpir pkgs/main/win-64::mpir-3.0.0-hec2e145_1mpmath pkgs/main/win-64::mpmath-1.3.0-py311haa95532_0
done
(chat_kimi) PS I:\chat_kimi\ChatTTS> python examples\web\webui.py
[+0800 20241014 05:22:44] [WARN] WebUI | funcs | no ffmpeg installed, use wav file output
[+0800 20241014 05:22:44] [INFO] WebUI | webui | loading ChatTTS model...
[+0800 20241014 05:22:44] [INFO] ChatTTS | dl | checking assets...
[+0800 20241014 05:22:45] [INFO] ChatTTS | dl | all assets are already latest.
[+0800 20241014 05:22:45] [INFO] ChatTTS | core | use device cuda:0
[+0800 20241014 05:22:45] [INFO] ChatTTS | core | vocos loaded.
[+0800 20241014 05:22:45] [INFO] ChatTTS | core | dvae loaded.
[+0800 20241014 05:22:46] [INFO] ChatTTS | core | embed loaded.
[+0800 20241014 05:22:47] [INFO] ChatTTS | core | gpt loaded.
[+0800 20241014 05:22:47] [INFO] ChatTTS | core | speaker loaded.
[+0800 20241014 05:22:47] [INFO] ChatTTS | core | decoder loaded.
[+0800 20241014 05:22:47] [INFO] ChatTTS | core | tokenizer loaded.
[+0800 20241014 05:22:47] [WARN] WebUI | funcs | Package nemo_text_processing not found!
[+0800 20241014 05:22:47] [WARN] WebUI | funcs | Run: conda install -c conda-forge pynini=2.1.5 && pip install nemo_text_processing
[+0800 20241014 05:22:47] [WARN] WebUI | funcs | Package WeTextProcessing not found!
[+0800 20241014 05:22:47] [WARN] WebUI | funcs | Run: conda install -c conda-forge pynini=2.1.5 && pip install WeTextProcessing
[+0800 20241014 05:22:47] [INFO] WebUI | webui | Models loaded successfully.
* Running on local URL: http://0.0.0.0:8080To create a public link, set `share=True` in `launch()`.这速度，倍之。
* Running on local URL: http://0.0.0.0:8080To create a public link, set `share=True` in `launch()`.
text: 0%| | 0/384(max) [00:00, ?it/s]C:\Users\dell\.conda\envs\chat_kimi\Lib\site-packages\transformers\models\llama\modeling_llama.py:655: UserWarning: 1Torch was not compiled with flash attention. (Triggered internally at C:\cb\pytorch_1000000000000\work\aten\src\ATen\native\transformers\cuda\sdp_utils.cpp:555.)attn_output = torch.nn.functional.scaled_dot_product_attention(
text: 0%|▏ | 1/384(max) [00:00, 3.58it/s]We detected that you are passing `past_key_values` as a tuple of tuples. This is deprecated and will be removed in v4.47. Please convert your cache or use an appropriate `Cache` class (https://huggingface.co/docs/transformers/kv_cache#legacy-cache-format)
text: 19%|██████████████▍ | 73/384(max) [00:02, 32.76it/s]
code: 25%|██████████████████▋ | 517/2048(max) [00:12, 42.36it/s]
看下时间,10秒音频，13秒完成。还可以。
下一节：从kimi中获取文本并，出声。