Skip to content

[TSL] Bump ml_dtypes to 0.5.0#17230

Closed
apivovarov wants to merge 1 commit intoopenxla:mainfrom
apivovarov:ml_dtypes_050
Closed

[TSL] Bump ml_dtypes to 0.5.0#17230
apivovarov wants to merge 1 commit intoopenxla:mainfrom
apivovarov:ml_dtypes_050

Conversation

@apivovarov
Copy link
Copy Markdown
Contributor

ml_dtypes 0.5.0. updates:

  • Added new 8-bit float types following IEEE 754 convention: ml_dtypes.float8_e4m3, ml_dtypes.float8_e3m4
  • Added the 8-bit floating point type ml_dtypes.float8_e8m0fnu, which is the OpenCompute MX scale format.
  • Added new 4-bit and 6-bit float types: ml_dtypes.float4_e2m1fn, ml_dtypes.float6_e2m3fn and ml_dtypes.float6_e3m2fn.
  • Fix outputs of float divmod and floor_divide when denominator is zero.

ml_dtypes/releases

Related PRs

- Added new 8-bit float types following IEEE 754 convention: ml_dtypes.float8_e4m3, ml_dtypes.float8_e3m4
- Added the 8-bit floating point type ml_dtypes.float8_e8m0fnu, which is the OpenCompute MX scale format.
- Added new 4-bit and 6-bit float types: ml_dtypes.float4_e2m1fn, ml_dtypes.float6_e2m3fn and ml_dtypes.float6_e3m2fn.
- Fix outputs of float divmod and floor_divide when denominator is zero.
@thomasjoerg
Copy link
Copy Markdown
Member

@ddunl Can you take a look at this PR please?

@ddunl
Copy link
Copy Markdown
Member

ddunl commented Nov 7, 2024

This was done in b36a74a so closing now

@ddunl ddunl closed this Nov 7, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants