Learned Spectral Compression
Compress the parameters of any PyTorch model by a factor of 10 without reducing accuracy. Just call `spectral(your_model_here)` and fine-tune your model with an additional compression loss term.
Compress the parameters of any PyTorch model by a factor of 10 without reducing accuracy. Just call `spectral(your_model_here)` and fine-tune your model with an additional compression loss term.
Log in or sign up for Devpost to join the conversation.