Deep Learning Model Explorer & Workbench Benchmark step stuck

[Mark_Rosenbaum]

Hi,

It seems that when the Benchmark step in the Deep Learning Model Explorer & Workbench is reached the process gets stuck.  Would it be possible to get some help with this?

-Mark

[Mark_Rosenbaum]

For extra reference my DevCloud ID is u218086 and I'm using DevCloud Edge.

[Mark_Rosenbaum]

On further investigation I found a bit more weirdness. For some reason I have two different UIDs showing in different places.(See Attached Screenshots)

[Zulkifli_Intel]

Hi Mark_Rosenbaum,

Thank you for reaching out.

You get 2 different UIDs because Intel Developer Cloud for the Edge and Intel® Tiber™ Developer Cloud for oneAPI are two separate development sandboxes.

For your issue with being stuck at the benchmark level, can you launch the JupyterLab?

For further investigation, please share with us these details :

1. Which Open Model Zoo model you selected
2. Model precision
3. Target environment and target architecture
4. Optimization Algorithm

Regards,

Zul

[Mark_Rosenbaum]

Hi Zul,

Thanks for the clarification about the UIDs, I had assumed they were universal.  As for JupyterLab I had made sure it worked before starting the benchmark but I'll go verify it works again just in case.

The following  model details were used for the benchmark:

It was the Sphereface model with the usage of  face_recognition and the precision set to FP16.  

It was run on the D-2712T Processor with the Target Architecture set to "AUTO"

The  Desired Optimization Algorithm was set to Latency Optimization

The following was the configuration displayed on the Benchmark tab:

Benchmark

Model:

Sphereface

Desired Precision:

FP16

Test Dataset Generation:

Number of test datasets will be auto-generated with reference to the input provided in the previous step

Target Environment:

AUTO

[Mark_Rosenbaum]

Here are screenshots from both dlworkbench and JupyterLab.

[Zulkifli_Intel]

Hi Mark_Rosenbaum,

Thank you for sharing the details. We are investigating this issue and will get back to you soon. 

Regards,

Zul

[Zulkifli_Intel · Accepted solution]

Hi Mark_Rosenbaum,

We have identified a hardware issue with the node. We have fixed the issue, could you please try again and let us know the result?

Regards,

Zul

[Mark_Rosenbaum]

Hi @Zulkifli_Intel,

Works like a charm now, thanks for such a quick fix.  I'll go ahead and accept your answer as the solution.

Thanks again,

Mark 

[Zulkifli_Intel]

Hi Mark_Rosenbaum,

I'm glad to hear that the issue has been resolved. Therefore this thread will no longer be monitored since this issue has been resolved. If you need any additional information from Intel, please submit a new question.

Regards,

Zul