Could you please make GGUF version too?

#1
by HR1777 - opened

Please make the quantized version of this model. Q5_k_m and Q6_k please.

Okay, I’ll look into and try to have it done soon!

thank you so much. I am waiting for it :)

It will probably be a day to two for it to load and train, and save with the quantized version. But, it's currently going!

And thank you for your support!

Had to do a little bit of research on it and was easy to understand. Got that working within minutes, but i had to retrain the entire model twice because, the first time I didn't set it up correctly to save. I am hoping I did it correctly, as it's looking promising, and should have it by today!.... I'm hoping.

Thank you so much for that. I'm looking forward to it!

HR1777 changed discussion status to closed

https://huggingface.co/ayjays132/PhillnetLargeQuantized

you will have to downlad and load it from local for now

Thank you so much for that. Although i was expecting Q5_k_M version instead of .bin files :). You are the best!

You are welcome! Phillnet is custom built off a hybrid mixed of multiple frameworks and pytorch seemed best for saving. I'll look into other ways for you.

Sign up or log in to comment