-
Notifications
You must be signed in to change notification settings - Fork 32
-
Notifications
You must be signed in to change notification settings - Fork 32
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
the issue of parameter settings #29
Comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I want to retest the code using a model with fewer parameters, but I’ve noticed that the --num-iters setting affects the final perplexity. I currently believe that --num-iters is a parameter that controls the amount of data to be judged, but I’m not certain. I want to know how I should set this parameter. The project provides dense model testing and sparse model testing for models with 66b and 175b parameters. I am a bit confused about the setting of --num-iters, and I hope to get your help. How should I set this parameter for models like 1.3b and 6.7b?
The text was updated successfully, but these errors were encountered: