Skip to content
This repository has been archived by the owner on Oct 13, 2022. It is now read-only.

about error of asg_train.py #66

Open
shanguanma opened this issue Jan 4, 2021 · 5 comments
Open

about error of asg_train.py #66

shanguanma opened this issue Jan 4, 2021 · 5 comments

Comments

@shanguanma
Copy link

I follow librispeech/asr/simple_v1/ recipe example, and rebuild k2 newest master(2021-1-4).
When I run python3 asg_train.py , It gives the following error:

scores: tensor([ 1.9269,  1.4873,  0.9007,  ...,  0.0990, -1.0238, -1.6546])
scores shape is :  torch.Size([52899])
Traceback (most recent call last):
  File "asg_train.py", line 430, in <module>
    main()
  File "asg_train.py", line 297, in main
    P.set_scores_stochastic_(scores)
TypeError: set_scores_stochastic_() takes 1 positional argument but 2 were given

I think it is a bug.

@csukuangfj
Copy link
Collaborator

Please refer to
#62 (comment)

@csukuangfj
Copy link
Collaborator

csukuangfj commented Jan 4, 2021

Please use k2-fsa/k2#563
It has not been merged yet.


Please use the latest k2, i.e., its master branch.

@shanguanma
Copy link
Author

OK, thanks, I will try it.

@csukuangfj
Copy link
Collaborator

Could you please post your result here when the decoding is done?
Just want to make sure that ASG training can be reproduced.

@shanguanma
Copy link
Author

Could you please post your result here when the decoding is done?
Just want to make sure that ASG training can be reproduced.

Yes, of course. Once completed, I will summarize the results here.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants