|
|
01abb5af21
|
additional support for float8_e4m3fnuz and _e5m2fnuz (#115214)
Follow up to #107586.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115214
Approved by: https://github.com/peterbell10, https://github.com/malfet
|
2024-01-22 18:33:41 +00:00 |
|
|
|
b637fdc8b3
|
Revert "additional support for float8_e4m3fnuz and _e5m2fnuz (#115214)"
This reverts commit 74e13624998f2a4de29bce73a949d7f0339ec04e.
Reverted https://github.com/pytorch/pytorch/pull/115214 on behalf of https://github.com/PaliC due to breaking internal builds ([comment](https://github.com/pytorch/pytorch/pull/115214#issuecomment-1900815152))
|
2024-01-19 17:35:04 +00:00 |
|
|
|
74e1362499
|
additional support for float8_e4m3fnuz and _e5m2fnuz (#115214)
Follow up to #107586.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115214
Approved by: https://github.com/peterbell10
|
2024-01-19 00:50:18 +00:00 |
|
|
|
c83a897348
|
[torch/csrc/onnx] Use nested namespaces (1/N) (#113991)
Differential Revision: [D51439849](https://our.internmc.facebook.com/intern/diff/D51439849)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/113991
Approved by: https://github.com/ZainRizvi
|
2023-11-18 00:20:10 +00:00 |
|
|
|
bd1229477d
|
[ONNX] Add initial support for FP8 ONNX export (#107962)
This PR resurrects @tcherckez-nvidia's #106379 with changes to resolve conflicts against newer `main` and defines our own constants for the new ONNX types to [avoid breaking Meta's internal usage of an old ONNX](https://github.com/pytorch/pytorch/pull/106379#issuecomment-1675189340).
- `::torch::onnx::TensorProto_DataType_FLOAT8E4M3FN=17`
- `::torch::onnx::TensorProto_DataType_FLOAT8E5M2=19`
Pull Request resolved: https://github.com/pytorch/pytorch/pull/107962
Approved by: https://github.com/justinchuby, https://github.com/titaiwangms
|
2023-09-08 20:40:39 +00:00 |
|
