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

[FEAT] Train Kara model #4

Open
sconetto opened this issue Nov 25, 2019 · 6 comments
Open

[FEAT] Train Kara model #4

sconetto opened this issue Nov 25, 2019 · 6 comments
Assignees
Labels
enhancement New feature or request help wanted Extra attention is needed

Comments

@sconetto
Copy link
Collaborator

Define and build kara model.

Expected Behavior

A model for the AI that can be trained in any machine given the dataset and dependencies.

@sconetto sconetto added enhancement New feature or request help wanted Extra attention is needed labels Nov 25, 2019
@sconetto
Copy link
Collaborator Author

First evolution at the training model:

  1. ffc2c54

@victorcmoura
Copy link
Contributor

I have uploaded a model with an accuracy of 0.7. But its result is far from the ideal. Maybe, more computational resources are needed. It is taking too long to train on my laptop =(

@sconetto
Copy link
Collaborator Author

I have uploaded a model with an accuracy of 0.7. But its result is far from the ideal. Maybe, more computational resources are needed. It is taking too long to train on my laptop =(

Sure, the minimum epoch count and step count should take 67 hours :(
I will try with a lower count to see if the result gets a bit better

@sconetto
Copy link
Collaborator Author

Moved model output to organize the project, tested some new training on commit:
54d57ec

@sconetto
Copy link
Collaborator Author

Model trained with:

  • 100 epochs
  • 2500 images from dataset
  • 20 steps per epoch

Output images:
image
image
image
image
image

Model from latest training on commit 3557265

@sconetto
Copy link
Collaborator Author

Model trained with:

  • 100 epochs
  • 999 images from dataset
  • 2 steps per epoch

Output images:

image
image
image
image
image

Due to the poor quality of the AI the model was not commited.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request help wanted Extra attention is needed
Projects
None yet
Development

No branches or pull requests

2 participants