We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
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
Hi! This repo is great! Found a small issue that may have been missed:
Llama 3.1 uses smoothing when computing the inverse frequencies in RoPE (https://github.com/huggingface/transformers/blob/049682a5a63042f087fb45ff128bfe281b2ff98b/src/transformers/modeling_rope_utils.py#L310) - this may result in some small differences between the baseline and the equinox models for the 3.1 variants when using long-contexts.
The text was updated successfully, but these errors were encountered:
No branches or pull requests
Hi!
This repo is great! Found a small issue that may have been missed:
Llama 3.1 uses smoothing when computing the inverse frequencies in RoPE (https://github.com/huggingface/transformers/blob/049682a5a63042f087fb45ff128bfe281b2ff98b/src/transformers/modeling_rope_utils.py#L310) - this may result in some small differences between the baseline and the equinox models for the 3.1 variants when using long-contexts.
The text was updated successfully, but these errors were encountered: