Facing missing positional argument : 'init_adadelta'

Alexander-H-Liu / End-to-end-ASR-Pytorch

This is an open source project (formerly named Listen, Attend and Spell - PyTorch Implementation) for end-to-end ASR implemented with Pytorch, the well known deep learning toolkit.

MIT License

1.18k stars 317 forks source link

Facing missing positional argument : 'init_adadelta' #53

Closed shamil-kadavan closed 4 years ago

shamil-kadavan commented 4 years ago

While testing the model on 'test-clean' dataset, i am facing issue "self.model = ASR(self.feat_dim, self.vocab_size, **self.config['model'])", this error is occurring because it is missing one required positional argument : 'inint_adadelta'

i only changed the max_steps, tf_steps and valid_step in asr_example.yaml config file

shamil-kadavan commented 4 years ago

Hi Team, the test_asr.py file needs some addition " init_adadelta = self.config['hparas']['optimizer'] == 'Adadelta'

    self.model = ASR(self.feat_dim, self.vocab_size,init_adadelta, **self.config['model'])

" This should be added in set_model in class Solver

Alexander-H-Liu commented 4 years ago

Hi @shamil-kadavan

you're right, we forgot to add this part in testing since the last update fixed in fa6683b60b57d1dc8fc27d16885f583752745d67 thanks

shamil-kadavan commented 4 years ago

Hi @Alexander-H-Liu , i am new to this Speech Recognition field. Can you help me with single file inference problem.