`pool_size` argument error for Global Average Pooling layer

Hi I’m trying to run a bi class classification on the following model shared as onnx model. It is supposed to work on classifying images as person and non person.
test_person224.zip (4.2 MB).
For the task I’m using now classification_tf2/byom_voc setup.
I’m able to run evaluation command on the TLTB model but unable to use it for training.
When I try to train the model in the notebook I get the following error:

I have attached a small sample of my dataset that can be used in case of replicating the issue. A sample specs file is also uploaded you can change it for replication.

Please let me know how to proceed further around this bug, so I can continue training and pruning this tltb model.

• Hardware 3090
• Network Type classification_tf2
• TLT Version nvidia-taov4.0.0
byom2-mobv2.tltb (7.9 MB)
mobilenet_pretr_v2_person_ep0.onnx (8.4 MB)
specs-train.yaml (1.2 KB)

Could you please share the full error log as well? Thanks.
I will check further.

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

Refer to Mobilenetv2 classification_tf1 pruning, error at pruned layer - #3 by Morganh to generate a new .tltb model.
I can run training successfully.

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