Spaces:
Running
on
T4
Running
on
T4
Update README.md
#29
by
hugginghead
- opened
README.md
CHANGED
@@ -27,7 +27,9 @@ You can simply duplicate the space to run it.
|
|
27 |
|
28 |
If running for the first time, create conda environment and install the desired torch version. Then install the rest of the requirements:
|
29 |
```
|
30 |
-
|
|
|
|
|
31 |
conda create --yes --name smlss_server python=3.8 libsndfile==1.0.31
|
32 |
conda activate smlss_server
|
33 |
conda install --yes pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
|
@@ -38,7 +40,7 @@ pip install -r requirements.txt
|
|
38 |
### Install frontend streaming-react-app dependencies
|
39 |
```
|
40 |
conda install -c conda-forge nodejs
|
41 |
-
cd streaming-react-app
|
42 |
npm install --global yarn
|
43 |
yarn
|
44 |
yarn build # this will create the dist/ folder
|
@@ -51,7 +53,7 @@ The server can be run locally with uvicorn below.
|
|
51 |
Run the server in dev mode:
|
52 |
|
53 |
```
|
54 |
-
cd seamless_server
|
55 |
uvicorn app_pubsub:app --reload --host localhost
|
56 |
```
|
57 |
|
|
|
27 |
|
28 |
If running for the first time, create conda environment and install the desired torch version. Then install the rest of the requirements:
|
29 |
```
|
30 |
+
git clone https://huggingface.co/spaces/facebook/seamless-streaming
|
31 |
+
cd seamless-streaming/seamless_server
|
32 |
+
conda config --append channels conda-forge
|
33 |
conda create --yes --name smlss_server python=3.8 libsndfile==1.0.31
|
34 |
conda activate smlss_server
|
35 |
conda install --yes pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
|
|
|
40 |
### Install frontend streaming-react-app dependencies
|
41 |
```
|
42 |
conda install -c conda-forge nodejs
|
43 |
+
cd ../streaming-react-app
|
44 |
npm install --global yarn
|
45 |
yarn
|
46 |
yarn build # this will create the dist/ folder
|
|
|
53 |
Run the server in dev mode:
|
54 |
|
55 |
```
|
56 |
+
cd ../seamless_server
|
57 |
uvicorn app_pubsub:app --reload --host localhost
|
58 |
```
|
59 |
|