UffParser: Unsupported number of graph 0

Accelerated Computing Intelligent Video Analytics TAO Toolkit
#1

0:00:10.152480524 5386 0x248db80 INFO nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger: NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1716> [UID = 1]: Trying to create engine from model files
ERROR: …/nvdsinfer/nvdsinfer_func_utils.cpp:33 [TRT]: UffParser: Unsupported number of graph 0
parseModel: Failed to parse UFF model
ERROR: tlt/tlt_decode.cpp:274 failed to build network since parsing model errors.
ERROR: …/nvdsinfer/nvdsinfer_model_builder.cpp:797 Failed to create network using custom network creation function
ERROR: …/nvdsinfer/nvdsinfer_model_builder.cpp:862 Failed to get cuda engine from custom library API
0:00:10.386728833 5386 0x248db80 ERROR nvinfer gstnvinfer.cpp:613:gst_nvinfer_logger: NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1736> [UID = 1]: build engine file failed
Segmentation fault (core dumped)

The above error exists even after giving the correct encode KEY.

#2

Can you share full command and config file?

#3

CONFIG FILE…
[property]
gpu-id=0
net-scale-factor=0.0039215697906911373
#0=RGB, 1=BGR
model-color-format=0

tlt-encoded-model=models/yolov4_cspdarknet_tiny_epoch_210_af.etlt
tlt-model-key=nvidia_tlt

int8-calib-file=models/

#model-engine-file=models/
labelfile-path=models/labels.txt

infer-dims=3;544;960
#uff-input-blob-name=Input
#output-blob-names=output_cov/Sigmoid;output_bbox/BiasAdd

maintain-aspect-ratio=1
uff-input-order=0
uff-input-blob-name=Input

0=FP32, 1=INT8, 2=FP16 mode

network-mode=0
num-detected-classes=7
gie-unique-id=1

0=Group Rectangles, 1=DBSCAN, 2=NMS, 3= DBSCAN+NMS Hybrid, 4 = None(No clustering)

cluster-mode=2

output-blob-names=BatchedNMS
parse-bbox-func-name=NvDsInferParseCustomBatchedNMSTLT
custom-lib-path=models/libnvds_infercustomparser.so
#maintain-aspect-ratio=0

[class-attrs-all]
nms-iou-threshold=0.5
pre-cluster-threshold=0.2
post-cluster-threshold=0.1
#group-threshold=1
eps=0.3
#detected-max-w=500

###########

#4

python3 deepstream_xavier_test_stable.py file:///opt/nvidia/deepstream/deepstream-5.1/samples/deepstream_python_apps/deepstream-nvdsanalytics/sample_videos/test.mp4

#5

Please check

  • if the key is correct
  • if models/yolov4_cspdarknet_tiny_epoch_210_af.etlt is available

Please follow GitHub - NVIDIA-AI-IOT/deepstream_tao_apps: Sample apps to demonstrate how to deploy models trained with TAO on DeepStream to run

#6

I have followed it, and the model is also present

#7

I am using DeepDtream-5.1, I have used 5.1 for TAO model earlier.
It is only the YOLO models causing the trouble.

#8

Please use the official yolo models and retry. Download it from the guide in the github.

#9

But why Official models, then there is no point in training custom models?

#10

This is to narrow down the issue.
If official model + key(nvidia_tlt) works,
but your mdoel + key(nvidia_tlt) does not work, there might be something mismatching for your model.

#11

Same Issue…

0:00:10.183842114 1194 0x3539180 INFO nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger: NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1716> [UID = 1]: Trying to create engine from model files
ERROR: …/nvdsinfer/nvdsinfer_func_utils.cpp:33 [TRT]: UffParser: Unsupported number of graph 0
parseModel: Failed to parse UFF model
ERROR: tlt/tlt_decode.cpp:274 failed to build network since parsing model errors.
ERROR: …/nvdsinfer/nvdsinfer_model_builder.cpp:797 Failed to create network using custom network creation function
ERROR: …/nvdsinfer/nvdsinfer_model_builder.cpp:862 Failed to get cuda engine from custom library API
0:00:10.450626540 1194 0x3539180 ERROR nvinfer gstnvinfer.cpp:613:gst_nvinfer_logger: NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1736> [UID = 1]: build engine file failed
Segmentation fault (core dumped)

#12

Resnet models are not causing any problem.
It is only the yolo models…

#13

Do you mean issue still happens when you run official yolo models?

#14

YES, the issue persists…

#15

It is not expected. Other users can run well with the official models.

One more thing needed to check, please check if there are additional characters after “nvidia_tlt”. Space characters are not expected.

#16

You mean other users can run with DeepStream-5.1?
And there is no way that there is a space, I have checked that too.

#17

Well, are there any changes to be made in the CONFIG FILE besides only the KEY?

#18

You can run with an old branch. GitHub - NVIDIA-AI-IOT/deepstream_tao_apps at release/tao3.0
Download official models and retry.

#19

Did that too… I am still facing the same error. How do I solve it?

#20

I share the searching result in the forum.
https://forums.developer.nvidia.com/search?expanded=true&q=parseModel:%20Failed%20to%20parse%20UFF%20model%20%20%23intelligent-video-analytics:tao-toolkit%20

Could you have a look to check if they can help you?