Star RTDETR
请先在RTDETR主页点击star以支持本项目
Star RTDETR to help more people discover this project.
Describe the bug
使用tensorRT 10.3导出engine时，报错
[08/30/2024-23:44:34] [I] [TRT] No checker registered for op: EfficientNMS_TRT. Attempting to check as plugin.
[08/30/2024-23:44:34] [I] [TRT] No importer registered for op: EfficientNMS_TRT. Attempting to import as plugin.
[08/30/2024-23:44:34] [I] [TRT] Searching for plugin: EfficientNMS_TRT, plugin_version: 1, plugin_namespace:
[08/30/2024-23:44:34] [I] [TRT] Successfully created plugin: EfficientNMS_TRT
[08/30/2024-23:44:34] [I] Finished parsing network model. Parse time: 0.16514
&&&& FAILED TensorRT.trtexec [TensorRT v100300] # trtexec --onnx=yolo_w_nms.onnx --saveEngine=rt_detr.engine --fp16
[08/30/2024-23:44:34] [W] [TRT] ModelImporter.cpp:420: Make sure input orig_target_sizes has Int64 binding.
[08/30/2024-23:44:34] [W] [TRT] onnxOpImporters.cpp:6119: Attribute class_agnostic not found in plugin node! Ensure that the plugin creator has a default value defined or the engine may fail to build.
[08/30/2024-23:44:34] [E] [TRT] ModelImporter.cpp:954: ERROR: ModelImporter.cpp:792 In function importModel:
[8] Assertion failed: (output_tensor_ptr->getType() != nvinfer1::DataType::kINT32 || output_trt_dtype == nvinfer1::DataType::kINT32) && "For INT32 tensors, the output type must also be INT32."
[08/30/2024-23:44:34] [E] Failed to parse onnx file
[08/30/2024-23:44:34] [E] Parsing model failed
[08/30/2024-23:44:34] [E] Failed to create engine from model or file.
[08/30/2024-23:44:34] [E] Engine set up failed
使用Netron检查时，最后输出维度含有？。
使用的是Custom dataset。 能使用训练后的pth->onnx->engine吗？
To Reproduce
latest.pth->onnx->engine
代码如下
from utils import yolo_insert_nms
yolo_insert_nms('D:/RT-DETR R18/model/model.onnx')
!trtexec --onnx=yolo_w_nms.onnx --saveEngine=rt_detr.engine --fp16
Star RTDETR 请先在RTDETR主页点击star以支持本项目 Star RTDETR to help more people discover this project.
Describe the bug 使用tensorRT 10.3导出engine时，报错 [08/30/2024-23:44:34] [I] [TRT] No checker registered for op: EfficientNMS_TRT. Attempting to check as plugin. [08/30/2024-23:44:34] [I] [TRT] No importer registered for op: EfficientNMS_TRT. Attempting to import as plugin. [08/30/2024-23:44:34] [I] [TRT] Searching for plugin: EfficientNMS_TRT, plugin_version: 1, plugin_namespace: [08/30/2024-23:44:34] [I] [TRT] Successfully created plugin: EfficientNMS_TRT [08/30/2024-23:44:34] [I] Finished parsing network model. Parse time: 0.16514 &&&& FAILED TensorRT.trtexec [TensorRT v100300] # trtexec --onnx=yolo_w_nms.onnx --saveEngine=rt_detr.engine --fp16 [08/30/2024-23:44:34] [W] [TRT] ModelImporter.cpp:420: Make sure input orig_target_sizes has Int64 binding. [08/30/2024-23:44:34] [W] [TRT] onnxOpImporters.cpp:6119: Attribute class_agnostic not found in plugin node! Ensure that the plugin creator has a default value defined or the engine may fail to build. [08/30/2024-23:44:34] [E] [TRT] ModelImporter.cpp:954: ERROR: ModelImporter.cpp:792 In function importModel: [8] Assertion failed: (output_tensor_ptr->getType() != nvinfer1::DataType::kINT32 || output_trt_dtype == nvinfer1::DataType::kINT32) && "For INT32 tensors, the output type must also be INT32." [08/30/2024-23:44:34] [E] Failed to parse onnx file [08/30/2024-23:44:34] [E] Parsing model failed [08/30/2024-23:44:34] [E] Failed to create engine from model or file. [08/30/2024-23:44:34] [E] Engine set up failed
使用Netron检查时，最后输出维度含有？。
使用的是Custom dataset。 能使用训练后的pth->onnx->engine吗？
To Reproduce
latest.pth->onnx->engine
代码如下
from utils import yolo_insert_nms
yolo_insert_nms('D:/RT-DETR R18/model/model.onnx')
!trtexec --onnx=yolo_w_nms.onnx --saveEngine=rt_detr.engine --fp16