torch.compile

torch._dynamo PyTorch’s compiler frontend which can output subgraphs when programs aren’t traceable. FX (Functional Transformations) FX is a toolkit for developers to use to transform nn.Module instances. FX consists of three main components: a symbolic tracer, an intermediate representation , and Python code generation . torch._inductor The default compiler in torch.compile, builds loops schedules and codegens triton or C++ kernels.

Topic	Replies	Views	Activity
About the torch.compile category torch.compile	0	1124	January 9, 2023
Hooks, dynamo, and unsupported builtins torch.compile	2	18	January 23, 2025
Aot_export_module captures more forward variables/parameters FX (Functional Transformations)	2	20	January 23, 2025
How to save torch.compile so we don't need to re-compile torch.compile	0	12	January 22, 2025
The difference between `torch.ops.aten.addmm.default` and `torch.ops.aten.linear.default` torch._dynamo	1	439	January 18, 2025
Does warming up the torch.compile necessarily trigger all cases? torch._dynamo	2	24	January 18, 2025
Does pytorch support building from source on arrch64 torch.compile	3	26	January 17, 2025
Torch._check failing with torch compile torch.compile	0	17	January 16, 2025
Graph breaks or No graph breaks in torchdynamo torch._dynamo	4	2379	January 14, 2025
How to get the optimized FX graph by TorchInductor programmatically? torch._inductor	0	17	January 13, 2025
Creating one giant C++ code covering code from all contexts torch._inductor	0	15	January 10, 2025
What is the uplimit of dynamo graph cache? torch._dynamo	2	18	January 10, 2025
Torch compile error when using flexattention + block mask torch.compile	0	32	January 8, 2025
Combine vmap, func.grad with tensordict FX (Functional Transformations)	0	29	January 7, 2025
Compiled_autograd Causes Abnormal Model Training Behavior torch.compile	1	19	January 6, 2025
Help understanding cpp codegen for inductor torch._inductor	5	64	January 2, 2025
`cuda.h' missing during torch.compile in new environment torch.compile	1	52	December 24, 2024
Does dynamo trigger real kernel execution? torch._dynamo	2	65	December 19, 2024
Compiling a model that uses sdp_kernel to enable the backends does not work torch._dynamo	4	45	December 18, 2024
Forward function not being compiled by default torch.compile	1	75	December 17, 2024
Onnx export with dynamo using torch.cond for dynamic models torch._dynamo	0	37	December 16, 2024
How to add custom operators to export_for_training? torch.compile	1	27	December 12, 2024
Sharing torch compile kernels between layers torch.compile	1	20	December 11, 2024
Is there a danger when doing DDP that different processes compiling kernels will overwrite each other? torch.compile	0	10	December 10, 2024
Saving modified fx graph torch.compile	0	65	December 9, 2024
Bug with isin assume_unique=True torch.compile	1	13	December 9, 2024
Scatter add much slower when compiled torch.compile	1	64	December 6, 2024
How to disable ___check_obj_id guards on dict? torch._dynamo	0	45	December 5, 2024
Not seeing any training time speedups when using torch.compile torch.compile	1	72	December 5, 2024
A node type in export IR graph FX (Functional Transformations)	1	52	November 28, 2024