You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2022/01/14 22:09:39 UTC

[GitHub] [tvm] masahi commented on pull request #9898: [Int8] Support cublas on e2e int8 models (also tried cudnn but doesn't work)

masahi commented on pull request #9898:
URL: https://github.com/apache/tvm/pull/9898#issuecomment-1013506772


   ping @Laurawly @Hzfengsy @comaniac what do you think? My conclusion is that cuDNN on NHWC + int8 is broken. Our test case at https://github.com/apache/tvm/blob/70de68a38f6738255cc71fd9a38964b23f2c5f55/tests/python/contrib/test_cudnn.py#L107 is also broken since it initializes input tensors with `np.random.uniform(-1, 1)` and all outputs are zero.
   
   But with this patch, we can at least compile quantized models having `qnn.dense` such as int8 BERT. 


-- 
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.

To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org