Morph MORPH
SIGN IN SIGN UP
dhruvhead / tt-mlir UNCLAIMED
0 0 0 MLIR

ACTIVITY

Commits Reviews
week month year
COMMITS
15
in the last week
CONTRIBUTORS
12
active
STARS
0
total
FORKS
0
total
TOP CONTRIBUTORS
V
Vincent Wells
4 commits
H
Het Shah
1 commit
J
Jack Zhang
1 commit
U
Uros Males
1 commit
D
Djordje Stefanovic
1 commit
A
Arjun Vangala
1 commit
A
Alex Baden
1 commit
P
Prem Kumar M
1 commit
J
Julia Grim
1 commit
A
Andjela Milovanovic
1 commit
RECENT COMMITS
H
[DSA] Add lowering support for the ttnn.experimental.indexer_score op (#8969)
Het Shah
97740f5
J
[TT-Lang] Turn TT-Lang ops back to DPS style and rematerialize custom_sharding_rule in RegisterCustomShardingRule (#9078)
Jack Zhang
724d2ff
U
Incrementally update EraseInverseOps const-trace cache (#8962)
Uros Males
6c71f4a
D
Model TTIR cache ops as in-place operations (#9091)
Djordje Stefanovic
6c43aaa
V
[D2M] Add d2m-jit mesh and global semaphore support (#9058)
Vincent Wells
99c621d
A
[d2m] arange blocking loop fix (#9114)
Arjun Vangala
a7b8eb8
V
[D2M] Reblock padded L1 outputs before DRAM transfer (#9057)
Vincent Wells
4f7ded7
V
[D2M] Preserve remote RAW ordering in DMA pipelining (#9050)
Vincent Wells
2315c59
V
[D2M] Recover flattened embedding index shape during lowering (#9088)
Vincent Wells
41f86d1
A
[d2m] Decompose SplitUnifiedThread into three passes; fix CB sync scoping (#8978)
Alex Baden
a6a60af
P
[TTNN] Fold Qwen3 QK-norm through decode RoPE fusing (fix #9110 perf, keep PCC) (#9125)
Prem Kumar M
55d1bae
J
[d2m-jit] Kernel autotuner (#9072)
Julia Grim
a48b1a1
A
[EmitPy] Fix multi-chip CPU-hoisted functions by capturing the device mesh shape (#9093)
Andjela Milovanovic
9478cd4
R
[TTNN] Remove legacy chain-based TTNNOptimizer path (#9086)
Radenko Pavlovic
fd1e8f1
V
Update test durations based on latest workflows 2026-07-27 (#9117)
Vladimir Milosevic
e9a7aea