[CUTLASS] Add blockwise scale gemm/bmm kernels (
#17789 )
Pull request merge
yongwwwpushed 1 commit to main • aaf185b…b0ccfb3 • 1 hour ago
[Relax][PyTorch] Support softshrink op for ExportedProgram (
#17786 )
Pull request merge
yongwwwpushed 1 commit to main • 3f16ec2…aaf185b • 6 hours ago
[Relax][PyTorch] Add support for where, cumprod and reciprocal ops (
#… Pull request merge
yongwwwpushed 1 commit to main • 81f7da8…3f16ec2 • 6 hours ago
[Relax][PyTorch] Support prod, std and var ops for ExportedProgram im…
[Relax][PyTorch] Support prod, std and var ops for ExportedProgram im…
Pull request merge
mshr-hpushed 1 commit to main • 95cbdaa…81f7da8 • 3 days ago
[Relax] Allow ingesting tensor.chunk() from exported torch program (
#… Pull request merge
[Relax] Enable bfloat16 for softmax struct-info inference (
#17781 )
[Relax] Enable bfloat16 for softmax struct-info inference (
#17781 )
Pull request merge
yongwwwpushed 1 commit to main • 26041f8…90391bb • 4 days ago
[Relax][Frontend] Support max/min in frontend op interface (
#17782 )
Pull request merge
tqchenpushed 1 commit to main • e60fd80…26041f8 • 5 days ago
[3rdparty] Enable bfloat16 for custom allreduce kernel (
#17780 )
Pull request merge
tqchenpushed 1 commit to main • 41c9c3b…e60fd80 • 5 days ago
[REFACTOR][TIR] remove legacy tir::any (
#17783 )
Pull request merge
yongwwwpushed 1 commit to main • c962198…41c9c3b • 5 days ago
[REFACTOR] Phase out StackVM (
#17784 )
Pull request merge
tqchenpushed 1 commit to main • 3da0738…c962198 • 5 days ago
[Relax][PyTorch] Support log2, log10 and log1p ops for ExportedProgra…
[Relax][PyTorch] Support log2, log10 and log1p ops for ExportedProgra…
Pull request merge
[Relax] Batch norm correctness on eval mode (
#17752 )
Pull request merge
[Relax] check for tensor_meta in exported_program_translator (
#17774 )
[Relax] check for tensor_meta in exported_program_translator (
#17774 )
Pull request merge
[Relax] Tensor.split with uneven tensors (
#17757 )
Pull request merge
[Relax][PyTorch] Add support for prod, std and var ops (
#17772 )
[Relax][PyTorch] Add support for prod, std and var ops (
#17772 )
Pull request merge
Hzfengsypushed 1 commit to main • e74ded2…adce5a4 • 9 days ago
[Relax][PyTorch] Add support for log2, log10 and log1p ops (
#17766 )
[Relax][PyTorch] Add support for log2, log10 and log1p ops (
#17766 )
Pull request merge
Hzfengsypushed 1 commit to main • a0d0859…e74ded2 • 11 days ago
[FIX][RELAX] fix fusion of transpose + matmul when constant weight (
#…
[FIX][RELAX] fix fusion of transpose + matmul when constant weight (
#… Pull request merge
Hzfengsypushed 1 commit to main • b0f1433…a0d0859 • 12 days ago
[Fix] Fix OpenCL header in attention utils (
#17762 )
[Fix] Fix OpenCL header in attention utils (
#17762 )
Pull request merge
[Relax][PyTorch] Add support for lerp, select and clone ops (
#17760 )
[Relax][PyTorch] Add support for lerp, select and clone ops (
#17760 )
Pull request merge
[Dlight] Fix general reduction rule to support non-last reduction axis (
Pull request merge
[Relax] Move TIR backend to gpu_generic (
#17749 )
You can’t perform that action at this time.