Delta-ML / delta

DELTA is a deep learning based natural language and speech processing platform.
https://delta-didi.readthedocs.io/
Apache License 2.0
1.59k stars 288 forks source link

When I run IEMOCAP ,I got an error #242

Closed Luyizhe closed 3 years ago

Luyizhe commented 3 years ago

When I bash run.sh,I got an error like below.

[ 2020-11-04 22:00:26,305 INFO api.py:541 41096 ] xxxx input: [], training: None Traceback (most recent call last): File "/opt1/luyz/DiDiDelta/delta/./delta/main.py", line 145, in entry() File "/opt1/luyz/DiDiDelta/delta/./delta/main.py", line 132, in entry app.run(main) File "/home/luyz/.conda/envs/py36/lib/python3.6/site-packages/absl/app.py", line 303, in run _run_main(main, args) File "/home/luyz/.conda/envs/py36/lib/python3.6/site-packages/absl/app.py", line 251, in _run_main sys.exit(main(argv)) File "/opt1/luyz/DiDiDelta/delta/./delta/main.py", line 93, in main solver.train_and_eval() File "/opt1/luyz/DiDiDelta/delta/delta/utils/solver/asr_solver.py", line 190, in train_and_eval self.model_fn(mode=utils.TRAIN) File "/opt1/luyz/DiDiDelta/delta/delta/utils/solver/keras_base_solver.py", line 140, in model_fn self.build(multi_gpu=(mode == utils.TRAIN)) File "/opt1/luyz/DiDiDelta/delta/delta/utils/solver/keras_base_solver.py", line 154, in build self.model, gpus=self.ngpu, cpu_relocation=False, cpu_merge=False) File "/home/luyz/.local/lib/python3.6/site-packages/tensorflow_core/python/util/deprecation.py", line 324, in new_func return func(*args, *kwargs) File "/home/luyz/.local/lib/python3.6/site-packages/tensorflow_core/python/keras/utils/multi_gpu_utils.py", line 240, in multi_gpu_model outputs = model(inputs) File "/home/luyz/.local/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/base_layer.py", line 847, in call outputs = call_fn(cast_inputs, args, **kwargs) File "/home/luyz/.local/lib/python3.6/site-packages/tensorflow_core/python/autograph/impl/api.py", line 237, in wrapper raise e.ag_error_metadata.to_exception(e) TypeError: in converted code:

/opt1/luyz/DiDiDelta/delta/delta/models/speech_cls_model.py:149 call  *
    x = inputs['inputs']

TypeError: list indices must be integers or slices, not str

Then I debug to identify the location of the error , I find that the variable 'ds' doesn't get any value after the code 'ds = input_fn()' in 'asr_solver.py'.

I think I have correct dataset dir ,or I can't run the code to stage 5 .Please give me some help.

Luyizhe commented 3 years ago

I got the problem above when I used configuration file “emo-keras-blstm.yml”。Today,I found if I use "emotion-speech-cls.yml" , I can complete training steps.

stale[bot] commented 3 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

stale[bot] commented 3 years ago

This issue is closed. Please re-open if needed.