AICVHub/Pytorch2TensorRT

Issues

• if args.dynamic: AttributeError: 'Namespace' object has no attribute 'dynamic'

  #15 opened 3 months ago by Jasonshao0123
  1

• TensorRT 9.X.0

  #14 opened 8 months ago by NevermindNilas
  0

• deleted

  #13 opened 9 months ago by lifan123-tech
  0

• 你好

  #10 opened a year ago by henbucuoshanghai
  3

• 怎样安装py版本 trt环境？pip?试了 不对

  #12 opened a year ago by henbucuoshanghai
  2

• engine转换失败

  #11 opened a year ago by Doctor-L-end
  0

• 怎么可视化推理结果

  #9 opened 2 years ago by ke-dev
  0

• 转int8模型时没有生成.cache文件

  #6 opened 4 years ago by hhhharold
  5

• do_inference报错

  #8 opened 4 years ago by evil-potato
  0

• 问题一：模型转为int8后，前处理需要相应的变化么，我看到你的calibrator里没有normalize这个操作。问题二：calibrator的batchsize和前推的batchsize是不是需要保持一致？

  #7 opened 4 years ago by LIUHAO121
  2

• 多输入输出分配内存

  #5 opened 4 years ago by budaLi
  4

• myCalibrator.py 中校验图片的 val.txt 的格式是什么呢

  #4 opened 4 years ago by forever208
  3

• 您好，想问一下批量推理测试图片应该怎么修改代码呢，期望您的回复，谢谢

  #3 opened 4 years ago by luluvi
  1

• 你好，我想用int8量化之后的模型部署，但是报错不支持int8，这个有什么方法可以解决吗？是版本不对还是跟显卡驱动有关

  #2 opened 4 years ago by young169
  3

• 推理不用c++部署？仍然用py？

  #1 opened 4 years ago by henbucuoshanghai
  1