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
Hello, I tried to train linear transformer with fp16 by DeepSpeed, and I got an error saying
RuntimeError: Expected object of scalar type Float but got scalar type Half for argument #2 'mat2' in call to _th_bmm_out
I edited line 67 in fast_transformers.attention.linear_attention to make it work
K = K * key_lengths.float_matrix[:, :, None, None].half()
Obviously, this is bad modification to do. How can I do when I get these kinds of errors?
The text was updated successfully, but these errors were encountered:
No branches or pull requests
Hello, I tried to train linear transformer with fp16 by DeepSpeed, and I got an error saying
I edited line 67 in fast_transformers.attention.linear_attention to make it work
K = K * key_lengths.float_matrix[:, :, None, None].half()
Obviously, this is bad modification to do.
How can I do when I get these kinds of errors?
The text was updated successfully, but these errors were encountered: