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.

Sign up for GitHub

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

Jump to bottom

slight difference from baichuan original alibi? #3

Open

DSXiangLi opened this issue Jan 31, 2024 · 0 comments

DSXiangLi commented Jan 31, 2024

powers = torch.arange(1, 1 + closest_power_of_2, dtype=torch.int32)

I think Baichuan starts from zero. Here is the orgiinal code from Baichuan

def _get_interleave_power_of_2(n):
start = 2 ** (-(2 ** -(math.log2(n) - 3)))
ratio = start
return [start * ratio ** i for i in range(n)]

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment