issues
search
cruise-automation
/
mlir-tcp
Tensor Compute Primitives: Mid-level Intermediate Representation for Machine Learning Programs
Other
35
stars
7
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
[NFC] Bump torch-mlir and torch pip deps
#99
sjain-stanford
closed
1 day ago
0
Add converter for index.Tensor_hacked_twin
#98
matthewfl
closed
1 week ago
0
Add support for integer division to TCP
#97
srinathava
closed
1 week ago
0
lack the TcpAttrs.td, is it unSupposed?
#96
GIN-Margarita
opened
3 weeks ago
1
add scatter to tcp custom op
#95
matthewfl
closed
3 weeks ago
0
relax aten.view conversion constraint
#94
zezhang
closed
4 weeks ago
0
support lowering for aten.log1p
#93
zezhang
closed
1 month ago
4
update custom op conversions
#92
zezhang
closed
1 month ago
0
[LLVM Integrate] Bump llvm, torch-mlir, stablehlo
#91
sjain-stanford
closed
1 month ago
1
fix broadcast condition
#90
zezhang
closed
1 month ago
0
fix broadcast condition
#89
zezhang
closed
1 month ago
2
Handle index_select by lowering it to tcp.gather
#88
navahgar
closed
1 month ago
0
Add support for customizing const inlining and ops with regions to IsolateGroupOps pass
#87
srinathava
closed
2 months ago
0
Slight improvement to fusion
#86
srinathava
closed
2 months ago
0
Revert multi-use fusion algorithm
#85
srinathava
closed
2 months ago
0
Revert multi use fusion algorithm because of correctness bug.
#84
srinathava
closed
2 months ago
0
update checking logic for scale and zp op
#83
zezhang
closed
2 months ago
0
Handle `tcp.bind_symbolic_shape` ops in fusion algorithm
#82
srinathava
closed
2 months ago
0
[AOT] Use torch dialect in backend form instead of raw torch dialect
#81
sjain-stanford
closed
2 months ago
0
Lower torch.aten.fake_quantize ops to tcp.custom_op
#80
zezhang
closed
2 months ago
0
[LLVM Integrate] Bump llvm and torch-mlir
#79
zezhang
closed
2 months ago
1
Symbolic shape expressions in TCP dialect
#78
sjain-stanford
closed
2 months ago
0
Tiling interface for slice op
#77
navahgar
closed
3 months ago
0
Cherry pick: Allow creation of TCP groups where an op has multiple uses
#76
srinathava
closed
3 months ago
0
[Clangd] Add TcpToTensor to clangd targets
#75
sjain-stanford
closed
3 months ago
0
Allow creation of TCP groups where an op has multiple uses
#74
srinathava
closed
3 months ago
0
[LLVM Integrate] Bump {stablehlo, torch-mlir, llvm}
#73
sjain-stanford
closed
3 months ago
1
Add tcp.slice op
#72
navahgar
closed
3 months ago
0
Remove index array type
#71
mabubakarpurdue
closed
4 months ago
1
[TorchToTcp] Mark torch ops with non-fp inputs dynamically legal
#70
sjain-stanford
closed
4 months ago
0
[NFC] Bump pip requirements
#69
sjain-stanford
closed
5 months ago
0
Add op for Gather
#68
navahgar
closed
5 months ago
1
[LLVM Integrate] Bump llvm, torch-mlir, stablehlo
#67
sjain-stanford
closed
5 months ago
0
Add ops for rounding
#66
navahgar
closed
5 months ago
0
[NFC] Project maintenance
#65
sjain-stanford
closed
5 months ago
1
[TorchToTcp] [AOT] Add test coverage for broadcast with rank increase
#64
sjain-stanford
closed
6 months ago
0
[TorchToTcp] [AOT] Lower `aten.expand` op to `tcp.broadcast`
#63
sjain-stanford
closed
6 months ago
0
[TorchToTcp] Don't broadcast axes that don't change size
#62
AaronStGeorge
closed
6 months ago
1
[T-REX] Add more elementwise ops to AOT test-suite and bugfix for `tcp.sqrt` with integer inputs
#61
sjain-stanford
closed
6 months ago
0
[AOT] Replace individual aot_compile definitions with starlark list comprehension
#60
sjain-stanford
closed
6 months ago
0
[NFC] Link to AOT readme from main readme
#59
sjain-stanford
closed
6 months ago
0
LIT test for `decompose-tensor-ops` pass
#58
sjain-stanford
closed
6 months ago
0
Adopt new dynamic_shapes API from torch.export
#57
sjain-stanford
closed
6 months ago
0
[AOT] Support `aten.add.Tensor` with `alpha`, and bufferize arith ops in llvm lowering
#56
sjain-stanford
closed
6 months ago
0
More legality fixes in dialect conversions + code consistency cleanup
#55
sjain-stanford
closed
6 months ago
0
TcpToLinalg legality fixes: Mark only the Tcp ops with conversions as illegal
#54
sjain-stanford
closed
6 months ago
0
[Lint] Add black formatter for python files
#53
sjain-stanford
closed
6 months ago
0
[AOT] Add tensor slicing to the aot_compile suite
#52
sjain-stanford
closed
6 months ago
0
[AOT] Add integer concat to the suite
#51
sjain-stanford
closed
6 months ago
0
Lower `aten.cat` -> `tensor::ConcatOp` and remove `tcp.concat`
#50
sjain-stanford
closed
6 months ago
0
Next