You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by "Johnson9009 (via GitHub)" <gi...@apache.org> on 2023/07/05 05:17:31 UTC
[GitHub] [tvm] Johnson9009 opened a new pull request, #15235: [QNN] Support Dequantize to "float16" and Quantize to "uint16"
Johnson9009 opened a new pull request, #15235:
URL: https://github.com/apache/tvm/pull/15235
Current the Relay qnn.dequantize lack of the support to "float16" output, only can dequantize a quant value to "float32", this PR add the lack function, so as "uint16" of qnn.quantize.
--
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
[GitHub] [tvm] Johnson9009 commented on a diff in pull request #15235: [QNN] Support Dequantize to "float16" and Quantize to "uint16"
Posted by "Johnson9009 (via GitHub)" <gi...@apache.org>.
Johnson9009 commented on code in PR #15235:
URL: https://github.com/apache/tvm/pull/15235#discussion_r1253262803
##########
python/tvm/relay/qnn/op/qnn.py:
##########
@@ -256,7 +256,7 @@ def simulated_quantize(data, output_scale, output_zero_point, axis=-1, out_dtype
return _make.simulated_quantize(data, out_dtype, output_scale, output_zero_point, axis)
-def dequantize(data, input_scale, input_zero_point, axis=-1):
+def dequantize(data, input_scale, input_zero_point, axis=-1, out_dtype="float32"):
r"""Dequantize op
This operator takes quantized int8 and unit8 as input and produces
Review Comment:
Done.
--
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
[GitHub] [tvm] Johnson9009 commented on pull request #15235: [QNN] Support Dequantize to "float16" and Quantize to "uint16"
Posted by "Johnson9009 (via GitHub)" <gi...@apache.org>.
Johnson9009 commented on PR #15235:
URL: https://github.com/apache/tvm/pull/15235#issuecomment-1624647331
@leandron @ibsidorenko @junrushao @tqchen @Hzfengsy
Is there anything need to change? If not, please help to check if we can merge it? Thanks.
--
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
[GitHub] [tvm] leandron commented on pull request #15235: [QNN] Support Dequantize to "float16" and Quantize to "uint16"
Posted by "leandron (via GitHub)" <gi...@apache.org>.
leandron commented on PR #15235:
URL: https://github.com/apache/tvm/pull/15235#issuecomment-1625085207
Thanks @Johnson9009 @ibsidorenko, this is merged now.
--
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
[GitHub] [tvm] tvm-bot commented on pull request #15235: [QNN] Support Dequantize to "float16" and Quantize to "uint16"
Posted by "tvm-bot (via GitHub)" <gi...@apache.org>.
tvm-bot commented on PR #15235:
URL: https://github.com/apache/tvm/pull/15235#issuecomment-1621041818
<!---bot-comment-->
Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from [Reviewers](https://github.com/apache/incubator-tvm/blob/master/CONTRIBUTORS.md#reviewers) by @-ing them in a comment.
<!--bot-comment-ccs-start-->
* cc @ibsidorenko <sub>See [#10317](https://github.com/apache/tvm/issues/10317) for details</sub><!--bot-comment-ccs-end-->
<sub>Generated by [tvm-bot](https://github.com/apache/tvm/blob/main/ci/README.md#github-actions)</sub>
--
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
[GitHub] [tvm] ibsidorenko commented on a diff in pull request #15235: [QNN] Support Dequantize to "float16" and Quantize to "uint16"
Posted by "ibsidorenko (via GitHub)" <gi...@apache.org>.
ibsidorenko commented on code in PR #15235:
URL: https://github.com/apache/tvm/pull/15235#discussion_r1252750608
##########
python/tvm/relay/qnn/op/qnn.py:
##########
@@ -256,7 +256,7 @@ def simulated_quantize(data, output_scale, output_zero_point, axis=-1, out_dtype
return _make.simulated_quantize(data, out_dtype, output_scale, output_zero_point, axis)
-def dequantize(data, input_scale, input_zero_point, axis=-1):
+def dequantize(data, input_scale, input_zero_point, axis=-1, out_dtype="float32"):
r"""Dequantize op
This operator takes quantized int8 and unit8 as input and produces
Review Comment:
Looks like this description is not correct anymore after all changes in input/output dtypes.
--
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
[GitHub] [tvm] leandron merged pull request #15235: [QNN] Support Dequantize to "float16" and Quantize to "uint16"
Posted by "leandron (via GitHub)" <gi...@apache.org>.
leandron merged PR #15235:
URL: https://github.com/apache/tvm/pull/15235
--
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