Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

MusicTransformer error: Non midi note detected. Only returning valid portion. Index, seed 5 (50, 2)0/100 00:02<00:00] #25

Open
BShennette opened this issue Nov 17, 2020 · 4 comments

Comments

@BShennette
Copy link

Hi, I was wondering if someone might have some insight.

I've followed the "Train" notebook, but when I try to run it, I get the titular error. I am using the .mid files from "/data/midi/examples" as training data or a different set.The "Index, seed #" number changes on different runs.

Additionally, I cannot run the Generate notebook with the example.pth file generated either, it gives the following:

\learner.py", line 18, in music_model_learner
if config is None: config = state['config']
KeyError: 'config'

@RoguePower
Copy link

This may resolve your error
encode_position = True
config = default_config()
config['encode_position'] = encode_position
learn_country = music_model_learner(data_country, config=config.copy())

@MatthewChau99
Copy link

I'm having the same error. Both "pred" and "full" contain non-midi notes if I train the model on my own. No error would happen if I use the pretrained model.

@MatthewChau99
Copy link

I found consecutive "xxsep"s in the predicted stream. Not sure what the reason is..

@MichaelLehenaur
Copy link

Hello @BShennette ,

could you resolve this issue with the anwers or by yourself?

I am currently facing the same problem (Non midi note detected. Only returning valid portion. Index, seed 5 (50, 2)) and i just can't find a propert solution or workaround...

Thanks in advance,
Michael

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

No branches or pull requests

4 participants