search

GuitarML / PedalNetRT

Deep Learning Networks for Real Time Guitar Effect Emulation using WaveNet with PyTorch
https://www.facebook.com/smartguitarml
GNU General Public License v3.0
349 stars 39 forks source link

Adopt Git LFS to optimise cloning speed #22

Open mishushakov opened 3 years ago

mishushakov commented 3 years ago

Cloning the repository takes a while

mish@ushakov Projects % git clone git@github.com:mishushakov/PedalNetRT.git
Cloning into 'PedalNetRT'...
remote: Enumerating objects: 68, done.
remote: Counting objects: 100% (68/68), done.
remote: Compressing objects: 100% (54/54), done.
remote: Total 278 (delta 28), reused 36 (delta 10), pack-reused 210
Receiving objects: 100% (278/278), 85.06 MiB | 1.48 MiB/s, done.
Resolving deltas: 100% (124/124), done.
Updating files: 100% (32/32), done.

consider adopting git-lfs to host the .wav samples remotely https://git-lfs.github.com

GuitarML commented 3 years ago

I've been thinking about this one, and you can test this with PedalNetRT, but for the SmartAmpPro, no changes are needed to the code to work with 16 bit PCM data. I'd be happy with swapping out the 32FP wav for 16PCM, and then each wav is only about 15MB. Sound quality seemed fine, and 16PCM is what CD quality is anyway. Just need to make sure the data is loaded as float32 numpy array.

mishushakov commented 3 years ago

Hey Keith, sure, but i was talking more about how it shouldn't be in version control

i can't make a pull request, because GitHub does not support adding files to Git LFS for forks

once you've got git-lfs, run

git lfs install
git lfs track "*.wav"
git add .gitattributes

what will happen next is the files will be uploaded to GitHub's LFS and the repository will only have symlinks instead of actual files

update: files from previous commits could also be migrated https://github.com/git-lfs/git-lfs/blob/main/docs/man/git-lfs-migrate.1.ronn

GuitarML commented 3 years ago

Yes that sounds like a good idea, the 16PCM would be in addition to using Lfs.