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

特征值量化的一点疑问 #4

Open theonly22 opened 4 years ago

theonly22 commented 4 years ago

class activation_quantize_fn(nn.Module): def init(self, a_bit): super(activation_quantize_fn, self).init() assert a_bit <= 8 or a_bit == 32 self.a_bit = a_bit self.uniform_q = uniform_quantize(k=a_bit)

def forward(self, x): if self.a_bit == 32: activation_q = x else: activation_q = self.uniform_q(torch.clamp(x, 0, 1))

print(np.unique(activation_q.detach().numpy()))

return activation_q

特征值量化前强行归一到（0，1），这一点不是很理解，有些特征值是比较大

666DZY666 commented 4 years ago

截断确实会造成误差，现在在截断前先进行了缩放以减小这个误差： class activation_quantize_fn(nn.Module): def init(self, a_bit): super(activation_quantize_fn, self).init() assert a_bit <= 8 or a_bit == 32 self.a_bit = a_bit self.uniform_q = uniform_quantize(k=a_bit)

def forward(self, x): if self.a_bit == 32: activation_q = x else: activation_q = self.uniform_q(torch.clamp(x 0.1, 0, 1)) #特征A截断前先进行缩放（ 0.1），以减小截断误差

print(np.unique(activation_q.detach().numpy()))

return activation_q

变换到（0，1）以使用通用的量化公式，误差是可以接受的