Update README.md
Browse files
README.md
CHANGED
@@ -85,30 +85,41 @@ It improves the quality of the lip-sync videos generated by the [Wav2Lip tool](h
|
|
85 |
|
86 |
1. Install [python 3.10.11](https://www.python.org/downloads/release/python-31011/)
|
87 |
2. Install [git](https://git-scm.com/downloads)
|
88 |
-
3.
|
|
|
|
|
89 |
```bash
|
90 |
python --version
|
91 |
git --version
|
|
|
92 |
```
|
93 |
Must return something like
|
94 |
```bash
|
95 |
Python 3.10.11
|
96 |
git version 2.35.1.windows.2
|
|
|
|
|
|
|
|
|
|
|
97 |
```
|
98 |
-
|
99 |
![Illustration](demo/visual_studio_1.png)
|
100 |
![Illustration](demo/visual_studio_2.png)
|
101 |
-
|
102 |
```bash
|
103 |
REM set PYTHON="your python.exe path"
|
104 |
```
|
105 |
-
|
106 |
set PYTHON="your python.exe path"
|
107 |
```
|
108 |
-
|
109 |
-
To do so, go to
|
|
|
|
|
|
|
110 |
![Illustration](demo/hf_aggrement.png)
|
111 |
-
|
112 |
1. Connect with your account
|
113 |
2. go to [access tokens](https://huggingface.co/settings/token) in settings
|
114 |
3. create a new token in read mode
|
@@ -119,7 +130,7 @@ To do so, go to [pyannote diarization-3.1 huggingface repository](https://huggin
|
|
119 |
"huggingface_token": "your token"
|
120 |
}
|
121 |
```
|
122 |
-
|
123 |
|
124 |
## Tutorial
|
125 |
- [FR version](https://youtu.be/43Q8YASkcUA)
|
|
|
85 |
|
86 |
1. Install [python 3.10.11](https://www.python.org/downloads/release/python-31011/)
|
87 |
2. Install [git](https://git-scm.com/downloads)
|
88 |
+
3. Install [Cuda 11.8](https://developer.nvidia.com/cuda-11-8-0-download-archive) if not ever done.
|
89 |
+
![Illustration](demo/cuda.png)
|
90 |
+
4. Check python and git installation
|
91 |
```bash
|
92 |
python --version
|
93 |
git --version
|
94 |
+
nvcc --version
|
95 |
```
|
96 |
Must return something like
|
97 |
```bash
|
98 |
Python 3.10.11
|
99 |
git version 2.35.1.windows.2
|
100 |
+
nvcc: NVIDIA (R) Cuda compiler driver
|
101 |
+
Copyright (c) 2005-2022 NVIDIA Corporation
|
102 |
+
Built on Wed_Sep_21_10:41:10_Pacific_Daylight_Time_2022
|
103 |
+
Cuda compilation tools, release 11.8, V11.8.89
|
104 |
+
Build cuda_11.8.r11.8/compiler.31833905_0
|
105 |
```
|
106 |
+
5. Install [Visual Studio](https://visualstudio.microsoft.com/fr/downloads/). During the install, make sure to include the Python and C++ packages in visual studio installer.
|
107 |
![Illustration](demo/visual_studio_1.png)
|
108 |
![Illustration](demo/visual_studio_2.png)
|
109 |
+
6. if you have multiple Python version on your computer edit wav2lip-studio.bat and change the following line:
|
110 |
```bash
|
111 |
REM set PYTHON="your python.exe path"
|
112 |
```
|
113 |
+
```bash
|
114 |
set PYTHON="your python.exe path"
|
115 |
```
|
116 |
+
7. pyannote.audio:You need to agree to share your contact information to access pyannote models.
|
117 |
+
To do so, go to :
|
118 |
+
- [pyannote diarization-3.1 huggingface repository](https://huggingface.co/pyannote/speaker-diarization-3.1)
|
119 |
+
- [pyannote segmentation-3.0 huggingface repository](https://huggingface.co/pyannote/segmentation-3.0)
|
120 |
+
set each field and click "Agree and access repository"
|
121 |
![Illustration](demo/hf_aggrement.png)
|
122 |
+
9. Create an access token to Huggingface:
|
123 |
1. Connect with your account
|
124 |
2. go to [access tokens](https://huggingface.co/settings/token) in settings
|
125 |
3. create a new token in read mode
|
|
|
130 |
"huggingface_token": "your token"
|
131 |
}
|
132 |
```
|
133 |
+
10. double click on wav2lip-studio.bat, that will install the requirements and download models
|
134 |
|
135 |
## Tutorial
|
136 |
- [FR version](https://youtu.be/43Q8YASkcUA)
|