In SONYCAudioClassifier config, the setting of both custom-lib-path and model-engine-file is necessary?

Hi there,

In config_infer_audio_sonyc.txt of SONYCAudioClassifier example, I’m trying to disable custom-lib-path, and only use model-engine-file, it runs successfully. So I’m curious about whether both custom-lib-path and model-engine-file are necessary when running this example.

[property]
gpu-id=0
net-scale-factor=1
onnx-file=./models/sonyc_audio_classify.onnx
model-engine-file=./models/sonyc_audio_classify.onnx_b2_gpu0_fp32.engine
labelfile-path=./models/audio_labels.txt
batch-size=2
## 0=FP32, 1=INT8, 2=FP16 mode
network-mode=0
num-detected-classes=31
gie-unique-id=1
output-blob-names=output1
network-type=1
#parse-classifier-func-name=NvDsInferParseCustomAudio
#custom-lib-path=./gstnvinferaudio_custom_parser/libnvds_infer_custom_parser_audio.so

Thanks.

I found it’s not necessary; the custom-lib-path example is a subcategory case.
In your case, if you have the requirement of subcategories classifiers, then you should assign custom-lib-path.

Please correct me if I’m wrong.

Thanks.

Yes,you are right.

https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_plugin_gst-nvinferaudio.html#id2

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.