Skip to content

[SpeechSynthesis/Tacotron2] Error while export ONNX models #487

Description

@LarryX

Related to SpeechSynthesis/Tacotron2
SpeechSynthesis/Tacotron2

Describe the bug
When running export Tacotron2 or WaveGlow ONNX models, it reports TypeError in load_and_setup_model()

root@32fdccde2ef1:/workspace/tacotron2# python exports/export_tacotron2_onnx.py --tacotron2 ./checkpoints/tacotron2_1032590_6000_amp -o output/ --fp16
Traceback (most recent call last):
  File "exports/export_tacotron2_onnx.py", line 405, in <module>
    main()
  File "exports/export_tacotron2_onnx.py", line 300, in main
    amp_run=args.fp16)
TypeError: load_and_setup_model() missing 1 required positional argument: 'cpu_run'

root@32fdccde2ef1:/workspace/tacotron2# python exports/export_waveglow_onnx.py --waveglow ./checkpoints/waveglow_1076430_14000_amp --wn-channels 256 -o output/ --fp16
Traceback (most recent call last):
  File "exports/export_waveglow_onnx.py", line 216, in <module>
    main()
  File "exports/export_waveglow_onnx.py", line 213, in main
    export_onnx(parser, args)
  File "exports/export_waveglow_onnx.py", line 168, in export_onnx
    args.fp16, forward_is_infer=False)
TypeError: load_and_setup_model() missing 1 required positional argument: 'cpu_run'

To Reproduce
Steps to reproduce the behavior:
Follow Instructions of:
https://github.com/NVIDIA/DeepLearningExamples/tree/master/PyTorch/SpeechSynthesis/Tacotron2/trt
and run Step #6

Expected behavior
No error while running the script.

Environment
Please provide at least:

  • Container version (e.g. pytorch:19.05-py3): pytorch:20.03-py3
  • GPUs in the system: (e.g. 8x Tesla V100-SXM2-16GB): Titian V
  • CUDA driver version (e.g. 418.67):440.33

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions