Guanghua Yu
Guanghua Yu
@HawkL327 可以参考PaddleSlim自动化压缩方案:https://github.com/PaddlePaddle/PaddleSlim/tree/develop/example/auto_compression/pytorch_yolov6 精度损失1个点左右,相比trt-fp16加速58%。
This repo is too old,No longer maintenance. please refer https://github.com/tianzhi0549/CTPN
是的!每张图片里一个字
因为类别较多,loss下降慢在合理情况内。
Dropout层可以加上试下,我没有加,在caffe上我迭代几千次已经明显收敛,batch size设为64
out_caffe_dir is the output path of dataset . The meaning of this sentence is to specify the output path
上述问题已修复~
@Fmaj7 onnx_format设为False,然后重新导出量化模型,预测时的报错信息可以发下吗?
> 看 报错信息应该具体是这几行,看起来是数据的dtype不匹配 > > ``` > ValueError: (InvalidArgument) The type of data we are trying to retrieve does not match the type of data currently contained in the container....
quantize_linear 这个算子是在onnx_format=True下出现的,你需要将dtype设置为int32,同时onnx_format=False