I have a lenet.onnx file that I can convert to onnx model using the python convert script (following) successfully, and then copy the files on the khadas board and execute it.
The parameters used to python script are:
#!/bin/bash
CONVERT_PATH="/acuity-toolkit/python/"
cd $CONVERT_PATH
./convert \
--model-name lenet \
--platform onnx \
--outputs "output" \
--inputs "input" \
--input-size-list '28,28,3' \
--model ./network/lenet.onnx \
--mean-values '127.5,127.5,127.5,255.0' \
--quantized-dtype asymmetric_affine \
--qtype int16 \
--kboard VIM3 \
--print-level 1
When I try to use the bash scripts to convert lenet onnx model, the shell scripts give different errors. I have been trying different parameters for all the three scripts i.e., but can then convert the model successfully.
0_import_model.sh
#!/bin/bash
NAME=lenet
ACUITY_PATH=../bin/
convert_onnx=${ACUITY_PATH}convertonnx
$convert_onnx \
--onnx-model ./model/lenet.onnx \
--net-output ${NAME}.json \
--data-output ${NAME}.data \
--outputs "output" \
--inputs "input" \
--input-size-list '3,28,28,1' \
1_quantize_model.sh
#!/bin/bash
NAME=lenet
ACUITY_PATH=../bin/
tensorzone=${ACUITY_PATH}tensorzonex
$tensorzone \
--action quantization \
--dtype int16 \
--channel-mean-value '127.5 127.5 127.5 255' \
--reorder-channel '2 1 0' \
--model-input ${NAME}.json \
--model-data ${NAME}.data \
--source text \
--source-file ./data/dataset0.txt \
--quantized-dtype dynamic_fixed_point-i16 \
--quantized-rebuild \
# --batch-size 4 \
# --input-size-list '28 28 3' \
The error snippet from the terminal looks like:
D Init coefficients ...
I Start tensor porvider ...
I Runing 1 epochs, algorithm: normal
I iterations: 0
Traceback (most recent call last):
File "tensorflow/python/client/session.py", line 1365, in _do_call
File "tensorflow/python/client/session.py", line 1350, in _run_fn
File "tensorflow/python/client/session.py", line 1443, in _call_tf_sessionrun
tensorflow.python.framework.errors_impl.OutOfRangeError: FIFOQueue '_0_fifo_queue' is closed and has insufficient elements (requested 1, current size 0)
[[{{node fifo_queue_DequeueMany}}]]
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "tensorzonex.py", line 445, in <module>
File "tensorzonex.py", line 386, in main
File "acuitylib/app/tensorzone/quantization.py", line 176, in run
File "acuitylib/app/tensorzone/quantization.py", line 154, in _run_quantization
File "acuitylib/app/tensorzone/graph.py", line 98, in run
File "tensorflow/python/client/session.py", line 958, in run
File "tensorflow/python/client/session.py", line 1181, in _run
File "tensorflow/python/client/session.py", line 1359, in _do_run
File "tensorflow/python/client/session.py", line 1384, in _do_call
tensorflow.python.framework.errors_impl.OutOfRangeError: FIFOQueue '_0_fifo_queue' is closed and has insufficient elements (requested 1, current size 0)
[[node fifo_queue_DequeueMany (defined at tensorflow/python/framework/ops.py:1949) ]]
Original stack trace for 'fifo_queue_DequeueMany':
File "tensorzonex.py", line 445, in <module>
File "tensorzonex.py", line 386, in main
File "acuitylib/app/tensorzone/quantization.py", line 176, in run
File "acuitylib/app/tensorzone/quantization.py", line 116, in _run_quantization
File "acuitylib/app/tensorzone/workspace.py", line 180, in _setup_graph
File "acuitylib/app/tensorzone/tensorprovider.py", line 159, in get_output
File "tensorflow/python/ops/data_flow_ops.py", line 488, in dequeue_many
File "tensorflow/python/ops/gen_data_flow_ops.py", line 3569, in queue_dequeue_many_v2
File "tensorflow/python/framework/op_def_library.py", line 744, in _apply_op_helper
File "tensorflow/python/framework/ops.py", line 3485, in _create_op_internal
File "tensorflow/python/framework/ops.py", line 1949, in __init__
[5650] Failed to execute script tensorzonex
Can you please guide what should be the right parameters for the shell scripts to convert a lenet onnx model?
@Frank @LukaChen @jasonl @numbqq @kenny