caffemodel to trt5 error

flipbecon · December 15, 2018, 6:52am

ERROR: …/builder/cudnnBuilder2.cpp (1508) - Misc Error in buildEngine: -1 (Could not find tensor ConvNd_1 in tensorScales.)
ERROR: …/builder/cudnnBuilder2.cpp (1508) - Misc Error in buildEngine: -1 (Could not find tensor ConvNd_1 in tensorScales.)

guanguojing1989 · December 17, 2018, 4:55am

Are you using int8 calibration ? I meet this problem when I’s using int8 calibration.
I solve this problem after I delete int8 calibration cache file generated by previous configuration.
Hope to help you.

flipbecon · December 24, 2018, 11:23am

thanks alot

Topic		Replies	Views
caffe2trt INT8 Quantization TensorRT	2	1374	October 12, 2021
Could not find tensor Relu in tensorScales INT8 Mode TensorRT TensorRT	0	338	August 5, 2019
Int8 calibrator issue TensorRT	0	456	October 23, 2019
RNNV2layer not support int8 Calibration in tensorrt4.0? TensorRT	1	715	February 27, 2019
Building INT8 inferencing engine of Caffe+Faster R-CNN meeted CudaError 4 in findFastestTactic. TensorRT	0	568	June 4, 2019
caffemodel to TensorRT problem TensorRT	2	992	December 9, 2019
Building INT8 engine fails: ../builder/cudnnBuilderWeightConverters.cpp:97:... TensorRT	2	795	October 12, 2021
Cask Error when using int8, fp32 works fine. TensorRT	1	454	March 9, 2020
Build Caffe with patch for TensorRT INT8 batch generation fails GPU-Accelerated Libraries	4	2087	January 17, 2018
Is there any method to build model with int8 weight in tensorrt? TensorRT	1	1225	July 29, 2021

caffemodel to trt5 error

Related topics