Skip to content

Commit

Permalink
readme
Browse files Browse the repository at this point in the history
  • Loading branch information
Joshua Meyer authored and WeberJulian committed Dec 13, 2023
1 parent 9cf6e83 commit ae5d0a8
Showing 1 changed file with 39 additions and 39 deletions.
78 changes: 39 additions & 39 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,70 +1,70 @@
# XTTS streaming server

## Running the server
## 1) Run the server

To run a pre-built container (CUDA 11.8):
### Recommended: use a pre-built container

CUDA 12.1:

```bash
$ docker run --gpus=all -e COQUI_TOS_AGREED=1 --rm -p 8000:80 ghcr.io/coqui-ai/xtts-streaming-server:latest
$ docker run --gpus=all -e COQUI_TOS_AGREED=1 --rm -p 8000:80 ghcr.io/coqui-ai/xtts-streaming-server:latest-cuda121
```

CUDA 12.1 version (for newer cards)
CUDA 11.8 (for older cards):

```bash
$ docker run --gpus=all -e COQUI_TOS_AGREED=1 --rm -p 8000:80 ghcr.io/coqui-ai/xtts-streaming-server:latest-cuda121
$ docker run --gpus=all -e COQUI_TOS_AGREED=1 --rm -p 8000:80 ghcr.io/coqui-ai/xtts-streaming-server:latest
```

Run with a custom XTTS v2 model (FT or previous versions):
Run with a fine-tuned model:

Make sure the model folder `/path/to/model/folder` contains the following files:
- `config.json`
- `model.pth`
- `vocab.json`

```bash
$ docker run -v /path/to/model/folder:/app/tts_models --gpus=all -e COQUI_TOS_AGREED=1 --rm -p 8000:80 ghcr.io/coqui-ai/xtts-streaming-server:latest`
```

Setting the `COQUI_TOS_AGREED` environment variable to `1` indicates you have read and agreed to
the terms of the [CPML license](https://coqui.ai/cpml).
## Not Recommended: Build the container yourself

(Fine-tuned XTTS models also are under the [CPML license](https://coqui.ai/cpml))
To build the Docker container Pytorch 2.1 and CUDA 11.8 :

## Testing the server

### Using the gradio demo
`DOCKERFILE` may be `Dockerfile`, `Dockerfile.cpu`, `Dockerfile.cuda121`, or your own custom Dockerfile.

```bash
$ python -m pip install -r test/requirements.txt
$ python demo.py
$ cd server
$ docker build -t xtts-stream . -f DOCKERFILE
$ docker run --gpus all -e COQUI_TOS_AGREED=1 --rm -p 8000:80 xtts-stream
```

### Using the test script
Setting the `COQUI_TOS_AGREED` environment variable to `1` indicates you have read and agreed to
the terms of the [CPML license](https://coqui.ai/cpml). (Fine-tuned XTTS models also are under the [CPML license](https://coqui.ai/cpml))

```bash
$ cd test
$ python -m pip install -r requirements.txt
$ python test_streaming.py
```
## 2) Testing the running server

## Building the container
Once your Docker container is running, you can test that it's working properly. You will need to run the following code from a fresh terminal.
1. To build the Docker container Pytorch 2.1 and CUDA 11.8 :
### Clone `xtts-streaming-server`
```bash
$ cd server
$ docker build -t xtts-stream .
```
For Pytorch 2.1 and CUDA 12.1 :
```bash
$ cd server
docker build -t xtts-stream . -f Dockerfile.cuda121
$ git clone [email protected]:coqui-ai/xtts-streaming-server.git
```
2. Run the server container:
### Using the gradio demo
```bash
$ docker run --gpus all -e COQUI_TOS_AGREED=1 --rm -p 8000:80 xtts-stream
$ cd xtts-streaming-server
$ python -m pip install -r test/requirements.txt
$ python demo.py
```
Setting the `COQUI_TOS_AGREED` environment variable to `1` indicates you have read and agreed to
the terms of the [CPML license](https://coqui.ai/cpml).


Make sure the model folder contains the following files:
- `config.json`
- `model.pth`
- `vocab.json`
### Using the test script
```bash
$ cd xtts-streaming-server
$ cd test
$ python -m pip install -r requirements.txt
$ python test_streaming.py
```

0 comments on commit ae5d0a8

Please sign in to comment.