Traceback (most recent call last):
File "hailo_model_zoo/main.py", line 178, in <module>
main()
File "hailo_model_zoo/main.py", line 174, in main
run(args)
File "hailo_model_zoo/main.py", line 163, in run
return handlers[args.command](args)
File "/home/giorgio/Project/hailo_model_zoo/hailo_model_zoo/main_driver.py", line 116, in compile
_ensure_quantized(runner, logger, args, network_info)
File "/home/giorgio/Project/hailo_model_zoo/hailo_model_zoo/main_driver.py", line 29, in _ensure_quantized
_ensure_parsed(runner, logger, network_info, args)
File "/home/giorgio/Project/hailo_model_zoo/hailo_model_zoo/main_driver.py", line 42, in _ensure_parsed
parse_model(runner, network_info, ckpt_path=args.ckpt_path, results_dir=args.results_dir, logger=logger,
File "/home/giorgio/Project/hailo_model_zoo/hailo_model_zoo/core/main_utils.py", line 134, in parse_model
add_nms_postprocess(
File "/home/giorgio/.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_client/tools/core_postprocess/core_postprocess_api.py", line 197, in add_nms_postprocess
new_hn, new_weights = create_nms_postprocess(runner.get_hn_model(),
File "/home/giorgio/.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_client/tools/core_postprocess/core_postprocess.py", line 33, in create_nms_postprocess
postproc_creator.prepare_hn_and_weights()
File "/home/giorgio/.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_client/tools/core_postprocess/nms_postprocess.py", line 423, in prepare_hn_and_weights
self._add_proposal_generator_layers()
File "/home/giorgio/.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_client/tools/core_postprocess/nms_postprocess.py", line 369, in _add_proposal_generator_layers
proposal_generator_layer.update_output_shapes()
File "/home/giorgio/.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_common/hailo_nn/hn_layers/layer.py", line 506, in update_output_shapes
output_shape = self._calc_output_shape()
File "/home/giorgio/.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_common/hailo_nn/hn_layers/proposal_generator.py", line 67, in _calc_output_shape
raise UnsupportedModelError("Invalid input shapes for layer {}.".format(self.name))
hailo_sdk_common.hailo_nn.exceptions.UnsupportedModelError: Invalid input shapes for layer ssd_mobilenet_v2/proposal_generator1.
I put a print inside /.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_common/hailo_nn/hn_layers/proposal_generator.py
Hi,
Thanks for addressing us regarding this issue.
It is hard to identify the issue without .hn file.
Please send us a ticket: https://hailo.ai/contact-customer-support/ with .hn file.
Thank you
Hi,
I'm trying to convert a mobilenetv2-ssd model trained using Tensorflow 1.15 object detection API on 1 class with 300x300x3 as input shape.
I tried to use
hailo_model_zoo/main.py
to convert this model and i modified the.yaml
and.json
files in the following way:-json
Running:
python hailo_model_zoo/main.py compile --ckpt my_model/model.ckpt --calib-path /my_images/test --yaml my_model/ssd_mobilenet_v2.yaml
I get the following error:
I put a print inside
/.virtualenvs/hailo_mz/lib/python3.8/site-packages/hailo_sdk_common/hailo_nn/hn_layers/proposal_generator.py
print(boxes_shape, classes_shape, anchors, classes)
and I get these outputs: [-1, 19, 19, 12] [-1, 19, 19, 3] 3 1 [-1, 10, 10, 12] [-1, 10, 10, 128] 3 42
I suspect it is due to something related to layers inside the .json file "bbox_decoders" but I'm not able to solve this error; how can I solve it?
Thank you