Latest torch.compile topics

Topic	Replies	Views	Activity
How to avoid check_object_id (ID_MATCH) on nn.Module objects? torch.compile	2	17	September 2, 2024
Understanding dynamic shapes and guards and when it does/does not cause graph breaks torch._dynamo	1	35	August 28, 2024
Difference between the graph break reasons: `Dynamic control flow is not supported at the moment.` and `generic_jump TensorVariable()` torch.compile	0	35	August 30, 2024
Unsupported: setattr when exporting the WavLM model torch.compile	0	18	August 29, 2024
How to enable `TORCH_USE_CUDA_DSA` torch.compile	8	12693	August 28, 2024
Custom Triton kernel with torch compile returns incorrect outputs torch.compile	3	54	August 28, 2024
Torch.fx.symbolic_trace with multiple GPUs FX (Functional Transformations)	0	9	August 27, 2024
AssertionError: torch.float32 != torch.bfloat16 torch.compile	0	14	August 27, 2024
How to gracefully mask CompositeImplicitAutograd for different backends torch._dynamo	0	26	August 26, 2024
How to turn a autograd.Function into custom_ops FX (Functional Transformations)	0	8	August 24, 2024
‘len’ is not supported in symbolic tracing by default FX (Functional Transformations)	7	2280	August 22, 2024
How to hint torch.export that a tensor is a constant? torch.compile	1	31	August 21, 2024
Torch compile error torch.compile	1	39	August 20, 2024
Error when export model to ONNX through dynamo_export torch.compile	4	1759	August 19, 2024
Torch.export + cudagraphs torch.compile	0	12	August 17, 2024
Torch_tensorrt compile for a range of sizes with steps torch._dynamo	1	21	August 17, 2024
Why core aten IR set doesn't contain scalar_tensor overload version of some ops, like bitwise-like ops, remainder and so on? torch.compile	0	18	August 14, 2024
Flex Attention return score values torch.compile	1	444	August 15, 2024
How to find specific "if ops" torch.compile	0	10	August 13, 2024
Fusing operators in torch.compile for Codegen torch._inductor	1	62	August 12, 2024
Recompiles on loss_kwargs changes torch.compile	1	10	August 10, 2024
How to make custom c++ class torch compilable torch.compile	0	24	August 9, 2024
What is the current/future best practice for custom autograd functions (with triton parts) torch.compile	0	39	August 9, 2024
How to Access Triton Kernels from TorchInductor when running on CPU? torch.compile	0	22	August 8, 2024
Torch.compile fails with C++ compile error: expected identifier before ‘[’ token torch.compile	4	104	August 7, 2024
Compiling stack of Conv3d increases runtime with redundant format conversions torch.compile	4	51	August 7, 2024
Is there any way to get traced GraphModule without creating torch dynamo's constraint guards? torch._dynamo	0	15	August 1, 2024
Add an input to a model with torch.compile FX (Functional Transformations)	0	17	July 31, 2024
Pytorch build fron source hangs with a few errors torch.compile	9	1154	July 30, 2024
Seeking Guidance on Integrating Custom Backends in Eager Mode with PyTorch torch.compile	0	14	July 29, 2024