mit-han-lab / smoothquant

[ICML 2023] SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models
https://arxiv.org/abs/2211.10438
MIT License
1.1k stars 127 forks source link

Question about code #76

Open Lucky-Lance opened 3 months ago

Lucky-Lance commented 3 months ago

Hi, I have a question about the code. The code here includes a view operation, but view is not an in-place operation in pytorch. An example is shown below, where I use 1.13.1+cu117.

image

I am confused. Would you please explain the utility of the view operation here? I think this line is unnecessary.