-
Notifications
You must be signed in to change notification settings - Fork 5.8k
Add batch_fc op in contrib #24017
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
Add batch_fc op in contrib #24017
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Does bias add and activation calculated separately?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
def batch_fc(input, param_size, param_attr, bias_size, bias_attr, act=None): | ||
""" | ||
**Batch FC layer** | ||
This Op can calculate BatcheFC. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
BatcheFC是拼写错误?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
OK, i will fix it.
def batch_fc(input, param_size, param_attr, bias_size, bias_attr, act=None): | ||
""" | ||
**Batch FC layer** | ||
This Op can calculate BatcheFC. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
也请解释下具体释义,目前不能直观理解op功能
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
OK, i will fix it,thanks~
This Op can calculate batch fc. Notice: It currently supports GPU device. This Op exists in contrib, which means that it is not shown to the public.
Part dim of batch_fc_op:
X.dim = slot_pairs_num * ins_num * in_dim
W.dim = slot_pairs_num * in_dim * out_dim
b.dim = slot_pairs_num * out_dim
output.dim = slot_pairs_num * ins_num * out_dim