Yizhou Jiang comments

Repositories
Issues
Comments

Results 2 comments of


                                            Yizhou Jiang

Question about the cross-antropy loss average?

thanks

Why this quantization model need more than 24GB GPU memory which is larger than ideal 500M?

> For me, it's not effective. Even with n_samples=1 on 32GB V100, it still leads to OOM. This is the launch script: python scripts/txt2img.py --prompt "a puppet wearing a hat"...