tonyskypc comments

Results 5 comments of


                                            tonyskypc

Question about INT8 quantization Slower

[yolov5s non-qat download](https://drive.google.com/file/d/17jIbbaUVt-0fzZLGJRfTApuuT7Ijar_D/view?usp=sharing) Thanks! @zerollzeng

Question about INT8 quantization Slower

@zerollzeng HI，Do as you say。 Trt8.4.1 **Non-QAT int8：** [07/18/2022-16:17:27] [I] === Performance summary === [07/18/2022-16:17:27] [I] Throughput: 827.176 qps [07/18/2022-16:17:27] [I] Latency: min = 2.11108 ms, max = 2.95532 ms,...

Question about INT8 quantization Slower

What you said is too professional. I can't operate it yet。 Please watch : https://github.com/maggiez0138/yolov5_quant_sample about qat code error？ Point out specific points [train_core_code and log download ](https://drive.google.com/file/d/1fDqAeWFIrcmOiEbf3hszoxeJunQNbHOq/view?usp=sharing) Thank you！！

Question about INT8 quantization Slower

[print each layer's inference time download](https://drive.google.com/file/d/1IzP4N3WxrqDeC5E_NALIppcn8MB5ZqeH/view?usp=sharing)，but I don't know how to solve it。 help me please，Thanks. @zerollzeng

RuntimeError: view size is not compatible with input tensor's size and stride (at least one dimension spans across two contiguous subspaces). Use .reshape(...) instead.

我将这个函数的所有view替换成reshape，重新编译安装，就可以了，你尝试一下。