Delete README.md
Browse files
README.md
DELETED
|
@@ -1,83 +0,0 @@
|
|
| 1 |
-
---
|
| 2 |
-
sdk: gradio
|
| 3 |
-
sdk_version: 5.32.1
|
| 4 |
-
---
|
| 5 |
-
# Whisper-WebUI
|
| 6 |
-
A Gradio-based browser interface for [Whisper](https://github.com/openai/whisper)
|
| 7 |
-
|
| 8 |
-
# Features
|
| 9 |
-
- Select the Whisper implementation you want to use between:
|
| 10 |
-
- [openai/whisper](https://github.com/openai/whisper)
|
| 11 |
-
- [SYSTRAN/faster-whisper](https://github.com/SYSTRAN/faster-whisper) (used by default)
|
| 12 |
-
- [Vaibhavs10/insanely-fast-whisper](https://github.com/Vaibhavs10/insanely-fast-whisper)
|
| 13 |
-
- Generate transcriptions from various sources, including **files** & **microphone**
|
| 14 |
-
- Currently supported output formats: **csv**, **srt** & **txt**
|
| 15 |
-
- Speech to Text Translation:
|
| 16 |
-
- From other languages to English (This is Whisper's end-to-end speech-to-text translation feature)
|
| 17 |
-
- Translate transcription files using Facebook NLLB models
|
| 18 |
-
- Pre-processing audio input with [Silero VAD](https://github.com/snakers4/silero-vad)
|
| 19 |
-
- Post-processing with speaker diarization using the [pyannote](https://huggingface.co/pyannote/speaker-diarization-3.1) model:
|
| 20 |
-
- To download the pyannote model, you need to have a Huggingface token and manually accept their terms in the pages below:
|
| 21 |
-
1. https://huggingface.co/pyannote/speaker-diarization-3.1
|
| 22 |
-
2. https://huggingface.co/pyannote/segmentation-3.0
|
| 23 |
-
|
| 24 |
-
# Installation and Running
|
| 25 |
-
|
| 26 |
-
- ## Run Locally
|
| 27 |
-
|
| 28 |
-
### Prerequisite
|
| 29 |
-
To run this WebUI, you need to have `git`, `python` version 3.8 ~ 3.10 & `FFmpeg`.<BR>If you're not using an Nvida GPU, or using a different `CUDA` version than 12.4, edit the file `requirements.txt` to match your environment.
|
| 30 |
-
|
| 31 |
-
Please follow the links below to install the necessary software:
|
| 32 |
-
- git : [https://git-scm.com/downloads](https://git-scm.com/downloads)
|
| 33 |
-
- python : [https://www.python.org/downloads/](https://www.python.org/downloads/)
|
| 34 |
-
- FFmpeg : [https://ffmpeg.org/download.html](https://ffmpeg.org/download.html)
|
| 35 |
-
- CUDA : [https://developer.nvidia.com/cuda-downloads](https://developer.nvidia.com/cuda-downloads)
|
| 36 |
-
|
| 37 |
-
After installing `FFmpeg`, make sure to **add** the `FFmpeg/bin` folder to your system `PATH`
|
| 38 |
-
|
| 39 |
-
### Installation using the script files
|
| 40 |
-
|
| 41 |
-
1. Download the the repository and extract its contents
|
| 42 |
-
2. Run `install.bat` or `install.sh` to install dependencies (It will create a `venv` directory and install dependencies there)
|
| 43 |
-
3. Start WebUI with `start-webui.bat` or `start-webui.sh` (It will run `python app.py` after activating the venv)
|
| 44 |
-
|
| 45 |
-
- ## Running with Docker
|
| 46 |
-
|
| 47 |
-
1. Install and launch [Docker-Desktop](https://www.docker.com/products/docker-desktop/)
|
| 48 |
-
|
| 49 |
-
2. Get the repository
|
| 50 |
-
|
| 51 |
-
3. If needed, update the `docker-compose.yaml` to match your environment
|
| 52 |
-
|
| 53 |
-
4. Docker commands:
|
| 54 |
-
|
| 55 |
-
Build the image ( Image is about ~7GB)
|
| 56 |
-
```sh
|
| 57 |
-
docker compose build
|
| 58 |
-
```
|
| 59 |
-
|
| 60 |
-
Run the container
|
| 61 |
-
```sh
|
| 62 |
-
docker compose up
|
| 63 |
-
```
|
| 64 |
-
|
| 65 |
-
5. Connect to the WebUI with your browser at `http://localhost:7860`
|
| 66 |
-
|
| 67 |
-
# VRAM Usages
|
| 68 |
-
- This project is integrated with [faster-whisper](https://github.com/guillaumekln/faster-whisper) by default for better VRAM usage and transcription speed.<BR>According to faster-whisper, the efficiency of the optimized whisper model is as follows:
|
| 69 |
-
| Implementation | Precision | Beam size | Time | Max. GPU memory | Max. CPU memory |
|
| 70 |
-
|-------------------|-----------|-----------|-------|-----------------|-----------------|
|
| 71 |
-
| openai/whisper | fp16 | 5 | 4m30s | 11325MB | 9439MB |
|
| 72 |
-
| faster-whisper | fp16 | 5 | 54s | 4755MB | 3244MB |
|
| 73 |
-
|
| 74 |
-
- Whisper's original VRAM usage table for available models:
|
| 75 |
-
| Size | Parameters | English-only model | Multilingual model | Required VRAM | Relative speed |
|
| 76 |
-
|:------:|:----------:|:------------------:|:------------------:|:-------------:|:--------------:|
|
| 77 |
-
| tiny | 39 M | `tiny.en` | `tiny` | ~1 GB | ~32x |
|
| 78 |
-
| base | 74 M | `base.en` | `base` | ~1 GB | ~16x |
|
| 79 |
-
| small | 244 M | `small.en` | `small` | ~2 GB | ~6x |
|
| 80 |
-
| medium | 769 M | `medium.en` | `medium` | ~5 GB | ~2x |
|
| 81 |
-
| large | 1550 M | N/A | `large` | ~10 GB | 1x |
|
| 82 |
-
|
| 83 |
-
Note: `.en` models are for English only, and you can use the `Translate to English` option from the other models
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|