opensource/DeOldify
3
0
Fork 0
mirror of https://github.com/jantic/DeOldify.git synced 2026-08-29 02:10:55 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #313 from jqueguiner/patch-1

Browse Source 
[DOC] Add model caching to doc
This commit is contained in:
Jason Antic
2021-02-01 16:32:19 -08:00
committed by GitHub
parent 1ef0db7925 b59a6a0038
commit 01b4c60f4c
1 changed files with 18 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+18
View File
@@ -380,6 +380,24 @@ curl -X POST "http://MY_SUPER_API_IP:5000/process" -H "accept: application/octet
```
> **Note:** If you don't have Nvidia Docker, [here](https://github.com/nvidia/nvidia-docker/wiki/Installation-(version-2.0)#installing-version-20) is the installation guide.
### Caching the model to improve API booting time
The API is made to download the model (if not already present locally) at boot time.
Adding the your model to the local subdirectory of the project "data/models" for instance :
- "/home/ubutun/deoldify/data/models/ColorizeArtistic_gen.pth" (image model)
- "/home/ubutun/deoldify/data/models/ColorizeVideo_gen.pth" (video model)
both models are available here:
- [Image Model](https://data.deepai.org/deoldify/ColorizeArtistic_gen.pth)
- [Video Model](https://data.deepai.org/deoldify/ColorizeVideo_gen.pth)
for ubuntu you could do :
```bash
$ wget -O /home/ubutun/deoldify/data/models/ColorizeArtistic_gen.pth https://data.deepai.org/deoldify/ColorizeArtistic_gen.pth
$ ## Then build the image
$ docker build -t api -f Dockerfile-api .
```
### Installation Details
This project is built around the wonderful Fast.AI library. Prereqs, in summary: