You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by an...@apache.org on 2022/11/07 17:39:37 UTC

[tvm] branch aluo/potential-turing-vortex-changes created (now 23d90bd41a)

This is an automated email from the ASF dual-hosted git repository.

andrewzhaoluo pushed a change to branch aluo/potential-turing-vortex-changes
in repository https://gitbox.apache.org/repos/asf/tvm.git


      at 23d90bd41a use builtin layernorm

This branch includes the following new commits:

     new 23d90bd41a use builtin layernorm

The 1 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails.  The revisions
listed as "add" were already present in the repository and have only
been added to this reference.



[tvm] 01/01: use builtin layernorm

Posted by an...@apache.org.
This is an automated email from the ASF dual-hosted git repository.

andrewzhaoluo pushed a commit to branch aluo/potential-turing-vortex-changes
in repository https://gitbox.apache.org/repos/asf/tvm.git

commit 23d90bd41aa3b9f71980ccbe674584bbd9fcd6b7
Author: Andrew Zhao Luo <an...@gmail.com>
AuthorDate: Mon Nov 7 09:39:23 2022 -0800

    use builtin layernorm
---
 python/tvm/relay/frontend/onnx.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/python/tvm/relay/frontend/onnx.py b/python/tvm/relay/frontend/onnx.py
index e244b4d9a1..393f7c56b6 100644
--- a/python/tvm/relay/frontend/onnx.py
+++ b/python/tvm/relay/frontend/onnx.py
@@ -1048,7 +1048,7 @@ class EmbedLayerNormalization(OnnxOpConverter):
         if segment_ids:
             vec_sum = _op.add(vec_sum, segment_vec)
 
-        ln = layer_norm(vec_sum, eps, gamma, beta)
+        ln = _op.nn.layer_norm(vec_sum, gamma, beta, epsilon=eps, axis=-1)
 
         mask_index = _op.const(np.zeros((batch_size,), dtype="int32"))
         if mask: