Incremental-Network-Quantization How can I reduce the model size?

How can I reduce the model size?

Open mychina75 opened this issue 7 years ago • 2 comments

Hi, Thank you for your amazing work. I tried to train my models, but the model size has no change. How can I reduce the model size, what else should I do?

thank you~

Sep 18 '17 03:09 mychina75

and in READM.md you mentioned "Real-time data shuffling is useful". Is this version support data shuffling operation?

Sep 18 '17 05:09 mychina75

hi, Q1: you can follow the paper "deep compression" or https://github.com/songhan/Deep-Compression-AlexNet.

Q2: if you load image directly, the caffe can shuffle data, or you can use https://github.com/lim0606/caffe-googlenet-bn

Sep 18 '17 14:09 AojunZhou

Incremental-Network-Quantization Incremental-Network-Quantization copied to clipboard

How can I reduce the model size?

Incremental-Network-Quantization
Incremental-Network-Quantization copied to clipboard