TAO classification export -- gen_ds_config does not work

Running classification export --gen_ds_config does not work, but the command does run successfully when the flag is omitted. Here is the output:

root@417b2b7ed9f3:/workspace# classification export -m pruned_output/weights/resnet_001.tlt -o model2.etlt -k my_key --data_type fp32 --gen_ds_config
Using TensorFlow backend.
Using TensorFlow backend.
WARNING:tensorflow:Deprecation warnings have been disabled. Set TF_ENABLE_DEPRECATION_WARNINGS=1 to re-enable them.
2021-10-04 18:44:33,722 [INFO] iva.common.export.keras_exporter: Using input nodes: ['input_1']
2021-10-04 18:44:33,723 [INFO] iva.common.export.keras_exporter: Using output nodes: ['predictions/Softmax']
NOTE: UFF has been tested with TensorFlow 1.14.0.
WARNING: The version of TensorFlow installed on this system is not guaranteed to work with UFF.
DEBUG: convert reshape to flatten node
DEBUG [/usr/local/lib/python3.6/dist-packages/uff/converters/tensorflow/converter.py:96] Marking ['predictions/Softmax'] as outputs
Traceback (most recent call last):
  File "/root/.cache/bazel/_bazel_root/ed34e6d125608f91724fda23656f1726/execroot/ai_infra/bazel-out/k8-fastbuild/bin/magnet/packages/iva/build_wheel.runfiles/ai_infra/iva/makenet/scripts/export.py", line 12, in <module>
  File "/root/.cache/bazel/_bazel_root/ed34e6d125608f91724fda23656f1726/execroot/ai_infra/bazel-out/k8-fastbuild/bin/magnet/packages/iva/build_wheel.runfiles/ai_infra/iva/common/export/app.py", line 233, in launch_export
  File "/root/.cache/bazel/_bazel_root/ed34e6d125608f91724fda23656f1726/execroot/ai_infra/bazel-out/k8-fastbuild/bin/magnet/packages/iva/build_wheel.runfiles/ai_infra/iva/common/export/app.py", line 215, in run_export
  File "/root/.cache/bazel/_bazel_root/ed34e6d125608f91724fda23656f1726/execroot/ai_infra/bazel-out/k8-fastbuild/bin/magnet/packages/iva/build_wheel.runfiles/ai_infra/iva/common/export/keras_exporter.py", line 389, in export
  File "/usr/lib/python3.6/os.py", line 220, in makedirs
    mkdir(name, mode)
FileNotFoundError: [Errno 2] No such file or directory: ''

Any guidance would be greatly appreciated.

Please try to modify -o model2.etlt to -o ./output/model2.etlt

Okay, that generated the nvinfer_config.txt but not labels.txt

It is strange. Could you share verbose log?

root@9eac581cb165:/workspace# classification export -m pruned_output/weights/resnet_001.tlt -o ./model.etlt --data_type fp32 -k mykey --gen_ds_config --verbose
Using TensorFlow backend.
Using TensorFlow backend.
WARNING:tensorflow:Deprecation warnings have been disabled. Set TF_ENABLE_DEPRECATION_WARNINGS=1 to re-enable them.
2021-10-06 17:23:36,072 [INFO] iva.common.export.keras_exporter: Using input nodes: ['input_1']
2021-10-06 17:23:36,072 [INFO] iva.common.export.keras_exporter: Using output nodes: ['predictions/Softmax']
2021-10-06 17:23:38,107 [DEBUG] iva.common.export.keras_exporter: Saving etlt model file at: ./model.etlt.
2021-10-06 17:23:39,963 [DEBUG] modulus.export._uff: Patching keras BatchNormalization...
2021-10-06 17:23:39,963 [DEBUG] modulus.export._uff: Patching keras Dropout...
2021-10-06 17:23:39,963 [DEBUG] modulus.export._uff: Patching UFF TensorFlow converter apply_fused_padding...
2021-10-06 17:23:40,679 [DEBUG] modulus.export._uff: Unpatching keras BatchNormalization layer...
2021-10-06 17:23:40,679 [DEBUG] modulus.export._uff: Unpatching keras Dropout layer...
NOTE: UFF has been tested with TensorFlow 1.14.0.
WARNING: The version of TensorFlow installed on this system is not guaranteed to work with UFF.
DEBUG: convert reshape to flatten node
DEBUG [/usr/local/lib/python3.6/dist-packages/uff/converters/tensorflow/converter.py:96] Marking ['predictions/Softmax'] as outputs
2021-10-06 17:23:42,259 [DEBUG] iva.common.export.base_exporter: Data file doesn't exist. Pulling input dimensions from the network.
2021-10-06 17:23:42,259 [DEBUG] iva.common.export.keras_exporter: Input dims: (3, 480, 720)

Confirmed that only nvinfer_config.txt is generated and not labels.txt

The label.txt is not generated. I’ll check further.

What docker image are you using? I could not use classification export.
I have the same problem with --gen_ds_config on Jupyter Notebook. It shows me the following error:
File “/usr/lib/python3.6/genericpath.py”, line 19, in exists
os.stat(path)
TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

@juanfeyero1
Latest tao 3.21.11 docker already fixes this issue.

For your new error, please create a new topic and also share the full command and full log.
Thanks.

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