-
Notifications
You must be signed in to change notification settings - Fork 3
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
Typo #15
Labels
bug
Something isn't working
Comments
You are right. it should be as you suggested. Can you open a PR? |
Just did over at the following link. This is my first pull request so please feel free to instruct me if I missed something. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
There is a typo in vit.jl where the following function is defined
function (m::MultiHeadAttention)(x::AbstractArray{T, 3}, ps, st) where {T}
More precisely the issue is within, where st_ is defined. I believe we want (minimally)
st_ = (qkv_layer=st_qkv, attention_dropout=st_attention, projection=st_projection)
where I've replaced attention -> attention_dropout. Otherwise Zygote.pullback gets confused when I try to run a training loop.
Apologies if it is taboo to not make a proper PR with mwe and error. Thanks again for all your work.
The text was updated successfully, but these errors were encountered: