-
Notifications
You must be signed in to change notification settings - Fork 90
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
[SYCLomatic] Add the bf16,fp16, float and double type floating point comparing tolerance. #1940
Conversation
Signed-off-by: Chen, Sheng S <[email protected]>
Signed-off-by: Chen, Sheng S <[email protected]>
Signed-off-by: Chen, Sheng S <[email protected]>
Signed-off-by: Chen, Sheng S <[email protected]>
Signed-off-by: Chen, Sheng S <[email protected]>
Signed-off-by: Chen, Sheng S <[email protected]>
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
default_tolerances = { | ||
"bf16": 9.77e-04, | ||
"fp16": 9.77e-04, | ||
"float": 1.19e-07, |
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.
for the end user, a little complex to set the float number here, how about change to set num of ULP instead.
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.
I prefer to keep this epsilon value here. As our floating-point comparison is developed by python script. And python code comparing the double type as the default type. Using the ULP to compare the different floating type need more work on it.
Signed-off-by: Chen, Sheng S <[email protected]>
Signed-off-by: Chen, Sheng S <[email protected]>
Signed-off-by: Chen, Sheng S <[email protected]>
No description provided.