search

666DZY666 / micronet

micronet, a model compression and deploy lib. compression: 1、quantization: quantization-aware-training(QAT), High-Bit(>2b)(DoReFa/Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference)、Low-Bit(≤2b)/Ternary and Binary(TWN/BNN/XNOR-Net); post-training-quantization(PTQ), 8-bit(tensorrt); 2、 pruning: normal、regular and group convolutional channel pruning; 3、 group convolution structure; 4、batch-normalization fuse for quantization. deploy: tensorrt, fp32/fp16/int8(ptq-calibration)、op-adapt(upsample)、dynamic_shape
MIT License
2.2k stars 478 forks source link

如何进行Dorefa的32位浮点型计算 #31

Open ghost opened 4 years ago

ghost commented 4 years ago

如果要做浮点型和八位量化的对比，32位浮点型就是把Wbits和Abits这两个参数设置为32，八位量化时就是设置为8吗？？

666DZY666 commented 3 years ago

是的。