gyshi commented on a change in pull request #15973: Numpy . implement numpy op
exp2 with tvm
URL: https://github.com/apache/incubator-mxnet/pull/15973#discussion_r318889459
##########
File path: contrib/tvmop/basic/ufunc.py
##########
@@ -98,3 +99,71 @@ def backward_vadd_gpu(dtype, ndim, reduce1st, req):
s[t].bind(bx, block_x)
s[t].bind(tx, thread_x)
return s, [X, in_grad_a, in_grad]
+
+def compute_exp2(dtype, ndim):
Review comment:
this day i am optimizing tvm op exp2, so i have no time to update this pr,
today i will updata it.
in my test, when input shape is vary large, the speed of tvm op exp2 is
faster than mxnet exp2
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
