- 26 Nov, 2023 1 commit
-
-
Umang Yadav authored
-
- 21 Nov, 2023 2 commits
-
-
Umang Yadav authored
-
Umang Yadav authored
-
- 20 Nov, 2023 9 commits
-
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
- 18 Nov, 2023 4 commits
-
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
- 17 Nov, 2023 16 commits
-
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
-
Umang Yadav authored
Handles all 4 Fp8 dtypes listed here : https://onnx.ai/onnx/technical/float8.html Follows saturation/clipping logic from table there as well : https://onnx.ai/onnx/technical/float8.html#cast Only adding fp8e4m3fnuz in MIGraphX IR for now.
-
Umang Yadav authored
-
Umang Yadav authored
-
- 16 Nov, 2023 1 commit
-
-
Artur Wojcik authored
-
- 14 Nov, 2023 1 commit
-
-
Umang Yadav authored
-
- 09 Nov, 2023 1 commit
-
-
Umang Yadav authored
-
- 08 Nov, 2023 2 commits
-
-
Zakor Gyula authored
The inaccuracy was caused by ONNX round requires nearest integer rounding for halway (0.5) cases. std::round rounds away from zero, thus giving wrong results with halfway cases. Replaced std::round with std::nearbyint which uses the correct rounding by default.
-
Brian Pickrell authored
-
- 07 Nov, 2023 2 commits
-
-
Paul Fultz II authored
-
Zakor Gyula authored
-
- 04 Nov, 2023 1 commit
-
-
Artur Wojcik authored
-
