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

Can't Train Feature Index #21

Open
iaminparis opened this issue Sep 21, 2023 · 9 comments
Open

Can't Train Feature Index #21

iaminparis opened this issue Sep 21, 2023 · 9 comments

Comments

@iaminparis
Copy link

Good afternoon, I ran into this error: (as a newbie, i just don't know)

Trying doing kmeans 397811 shape to 10k centers.

NameError Traceback (most recent call last)
in train_index(exp_dir1, version19)
32 big_npy = (
---> 33 MiniBatchKMeans(
34 n_clusters=10000,

NameError: name 'MiniBatchKMeans' is not defined

During handling of the above exception, another exception occurred:

NameError Traceback (most recent call last)
1 frames
in train_index(exp_dir1, version19)
42 )
43 except:
---> 44 info = traceback.format_exc()
45 logger.info(info)
46 infos.append(info)

NameError: name 'traceback' is not defined

@ardha27
Copy link
Owner

ardha27 commented Sep 21, 2023

which one of my code that you run? i don't remember we can edit the kmeans on my colabs

@iaminparis
Copy link
Author

TrainingV2_NoUI on Colab, is there a simpler method?

@iaminparis
Copy link
Author

Also, it crashes while training model (using too much RAM)

@iaminparis
Copy link
Author

missclicked

@iaminparis iaminparis reopened this Sep 21, 2023
@ardha27
Copy link
Owner

ardha27 commented Sep 21, 2023

can you provide screenshot. and how long is your dataset? do you split it up?

@iaminparis
Copy link
Author

my dataset might be a bit long, ~3 hours, split up every 5 seconds, let me try with less samples

@ardha27
Copy link
Owner

ardha27 commented Sep 21, 2023

thats might be the problem, i usually train with 5-15 minute dataset in colab. the longest dataset that i've been train is 40 minute

@iaminparis
Copy link
Author

Okay! Thank you, I'll tell you in a few minutes

@Sinvanor
Copy link

I've gotten the same error. I was able to get one model to work with around an hour, maybe a little more, but longer than that and it gives me this error. I have 569 5-10 second split audio files and the other that gave me this error has upwards of 920+. Is there anyway to make it so that it can handle more data? I don't mind if it takes longer to train and get epochs.

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

3 participants