onnx->trt F16:Subnormal FP16 value detected

 onnx->trt报的warning如下:

If this is not the desired behavior, please modify the weights or retrain with regularization to reduce the magnitude of the weights.
[10/12/2022-19:11:09] [TRT] [W] Weights [name=Conv_167 || Conv_170.weight] had the following issues when converted to FP16:
[10/12/2022-19:11:09] [TRT] [W]  - Subnormal FP16 values detected. 
[10/12/2022-19:11:09] [TRT] [W] If this is not the desired behavior, please modify the weights or retrain with regularization to reduce the magnitude of the weights.
[10/12/2022-19:11:09] [TRT] [W] Weights [name=Conv_167 || Conv_170.weight] had the following issues when converted to FP16:
[10/12/2022-19:11:09] [TRT] [W]  - Subnormal FP16 values detected. 
[10/12/2022-19:11:09] [TRT] [W] If this is not the desired behavior, please modify the weights or retrain with regularization to reduce the magnitude of the weights.
[10/12/2022-19:11:09] [TRT] [W] Weights [name=Conv_173 + PWN(PWN(Sigmoid_174), Mul_175).weight] had the following issues when converted to FP16:
[10/12/2022-19:11:09] [TRT] [W]  - Subnormal FP16 values detected. 
[10/12/2022-19:11:09] [TRT] [W] If this is not the desired behavior, please modify the weights or retrain with regularization to reduce the magnitude of the weights.
[10/12/2022-19:11:09] [TRT] [W] Weights [name=Conv_173 + PWN(PWN(Sigmoid_174), Mul_175).weight] had the following issues when converted to FP16:
[10/12/2022-19:11:09] [TRT] [W]  - Subnormal FP16 values detected. 
[10/12/2022-19:11:09] [TRT] [W] If this is not the desired behavior, please modify the weights or retrain with regularization to reduce the magnitude of the weights.
[10/12/2022-19:11:09] [TRT] [W] Weights [name=Conv_173 + PWN(PWN(Sigmoid_174), Mul_175).weight] had the following issues when converted to FP16:
[10/12/2022-19:11:09] [TRT] [W]  - Subnormal FP16 values detected. 

Subnrmal FP16 values detected - TensorRT - NVIDIA Developer Forums

英伟达官方答复:如果不影响精度可以忽略

Subnormal FP16 values detected - TensorRT - NVIDIA Developer Forums

  • 3
    点赞
  • 2
    收藏
    觉得还不错? 一键收藏
  • 3
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论 3
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值