quantization


Topic Replies Activity
About the quantization category 1 October 2, 2019
Help needed: Specific function call causing 20X slowdown of computation 3 February 25, 2020
Onnx export failed int8 model 8 February 25, 2020
The parameters saved in the checkpoint are different from the ones in the fused model 4 February 24, 2020
[caffe2] Post train quantization 2 February 24, 2020
Not able to load quantized model in android 2 February 24, 2020
How to use a quantized model on INT8 harware? 7 February 24, 2020
Current status of automatic quantization support 7 February 21, 2020
Decrease in the Speed of Quantization Model 4 February 20, 2020
Quantization support for 1D convolutions? 3 February 18, 2020
Conv2d_unpack and conv2d_prepack behavior 2 February 18, 2020
Quantized::cat running time is slower than fp32 model 4 February 17, 2020
RuntimeError: Unimplemented backend QuantizedCPU 2 February 14, 2020
AssertionError: torch.nn.quantized.ReLU does not support inplace 2 February 14, 2020
Quantized convolution and NHWC layout 3 February 14, 2020
RuntimeError: No function is registered for schema aten::thnn_conv2d_forward 9 February 14, 2020
When quantized::max_pool2d is used? 2 February 14, 2020
AssertionError: min nan should be less than max nan 4 February 14, 2020
Casting from 32b to 8 bit after accumulation in a multiplication 4 February 14, 2020
Pretrained quantized models' export to ONNX fails 5 February 13, 2020
8 bit quantization - modulo 256 inside convolutions? 1 February 4, 2020
Creat a tensor with random number of 1 and -1 3 February 1, 2020
How does PyTorch implement Quantization? 11 February 1, 2020
PyTorch 1.3 wheels for Raspberry Pi (Python 3.7) 6 January 31, 2020
Best way to quantize Transformer architecture 3 January 28, 2020
Can't load model after dynamic quantization 7 January 23, 2020
Issue with Quantization 1 January 23, 2020
How to extract individual weights after per channel static quantization? 1 January 23, 2020
Exception: must run observer before calling calculate_qparams! 2 January 21, 2020
The packing format of quantized parameters after jitting 2 January 20, 2020