| Fix some bugs and add FuseBatchMatMul pass.
| 23 天前 |
| Add two MatMul Fusion passes.
Please enter the commit message for your changes. Lines starting
| 23 天前 |
| Add some fusion passes.
| 23 天前 |
| Fix mfusion rank0 scalar and batch matmul lowering
Lower aclnn.batch_matmul transpose flags with explicit permutes before bmm/matmul conversion.
Use torch.prims.convert_element_type for mfuse.cast, rewrite splat vtensor literals through full, and cast mixed-type binary operands to the result element type.
Prevent DVM clustering for ops that consume unsafe external rank-zero tensor operands while allowing safe constant/full scalar producers.
| 23 天前 |
| Add some fusion passes.
| 23 天前 |
| Add MatMul Fusion passes in pipeline.
| 23 天前 |
| Add MatMul Fusion passes in pipeline.
| 23 天前 |
| fix: convert mfuse.reshape to torch.aten.reshape and disable fuse-matmul-reshape pass
- Convert mfuse.reshape to torch.aten.reshape instead of torch.aten.view
- Disable fuse-matmul-reshape pass from fusion pipeline
- The pass was ineffective because canonicalization folds reshapes with
identical input/output shapes
- Add test cases for reshape conversion
- Skip Python tests for disabled pass
| 23 天前 |
| Fixed permute problem.
| 23 天前 |
| Add MatMul Fusion passes in pipeline.
| 23 天前 |
| Fixed permute problem.
| 23 天前 |
| Fix gelu_backward decompose problem
| 23 天前 |
| Add Reduce post split pattern
| 23 天前 |
| refactor(mfusion): move RmsNorm fusion to Torch dialect pipeline
Move RmsNorm fusion from Mfuse Fusion to Torch Fusion before convert-torch-to-mfuse, so decomposed Torch RmsNorm chains can be fused earlier and lowered consistently. Remove legacy Mfuse-side FuseRmsNorm pass and adapt LIT/Python tests to Torch-side fusion semantics.
| 23 天前 |