-
Notifications
You must be signed in to change notification settings - Fork 720
Pull requests: llvm/torch-mlir
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[TorchToTosa] Fix aten.div.Tensor_mode overflow for i64 inputs
#4720
opened Aug 14, 2026 by
hariprasadravi
Contributor
Loading…
[Torch] Fix aten.linalg_vector_norm for ord = 0 and ord = +/-inf
#4718
opened Aug 13, 2026 by
hariprasadravi
Contributor
Loading…
[TorchToTosa] Add direct aten.addmm legalization
#4712
opened Aug 13, 2026 by
Tessil
Contributor
Loading…
Reject zero-output linear lowering to TOSA
#4688
opened Jul 27, 2026 by
mukul-sosalekeerthi-ARM
Collaborator
Loading…
[Torch] Use FP32 opmath for FP16 batch norm inference
#4683
opened Jul 24, 2026 by
gokcan-karakus
Collaborator
Loading…
Handle zero-extent rms_norm decomposition
#4678
opened Jul 23, 2026 by
mukul-sosalekeerthi-ARM
Collaborator
Loading…
[Torch] Fuse per-channel quantized convolution and matmul to integer ops
#4673
opened Jul 21, 2026 by
dimp-pl
Loading…
[Torch] Add FP32 addbmm decomposition
#4669
opened Jul 20, 2026 by
gokcan-karakus
Collaborator
Loading…
[TORCH] Add support for aten.index_add Op via torch decomposition
#4653
opened Jul 15, 2026 by
AiyyappanMR
Contributor
Loading…
[TorchOnnxToTorch] Make onnx.LayerNormalization bias (B) input optional
#4652
opened Jul 14, 2026 by
nchristo-synaptics
Loading…
support torch.transformers.grouped_mm_fallback
#4649
opened Jul 13, 2026 by
PlatinumCD
Contributor
Loading…
[TorchOnnxToTorch] Fix QLinearGlobalAveragePool kernel size for unknown result dims
#4646
opened Jul 10, 2026 by
dimp-pl
Loading…
[TorchOnnxToTorch] Dequantize the bias in the per-channel QLinearConv lowering
#4645
opened Jul 10, 2026 by
dimp-pl
Loading…
[TORCH] Add support for aten.smooth_l1 loss via torch decomposition
#4642
opened Jul 9, 2026 by
AiyyappanMR
Contributor
Loading…
[TorchToLinalg] Support k-dim loss in AtenNllLossForwardOp lowering
#4641
opened Jul 9, 2026 by
AkunCube
Loading…
[TorchOnnxToTorch] Fix onnx.Unique lowering for optional trailing outputs
#4640
opened Jul 8, 2026 by
javierdejesusda
Contributor
Loading…
[Torch] Restrict aten.ones/zeros/full folding to value tensors
#4636
opened Jul 7, 2026 by
gaoweifan
Loading…
[TORCH] Add support for aten.addbmm op via torch decomposition
#4635
opened Jul 7, 2026 by
AiyyappanMR
Contributor
Loading…
[TORCH] Add support for aten.histc op via torch decomposition
#4632
opened Jul 6, 2026 by
AiyyappanMR
Contributor
Loading…
Previous Next
ProTip!
Adding no:label will show everything without a label.