|
About the FX (Functional Transformations) category
|
|
0
|
1266
|
February 25, 2021
|
|
"Offset increment outside graph capture encountered unexpectedly" is a specific PyTorch Windows issue that's known to occur with CUDA Graphs. This is a bug in PyTorch Windows builds
|
|
0
|
69
|
February 10, 2026
|
|
How to verify that default_decompositions successfully reduce operators to the Core ATen IR set?
|
|
0
|
16
|
December 30, 2025
|
|
TS2EPConverter and `dynamic_shapes`
|
|
0
|
91
|
October 5, 2025
|
|
Autofunctionalize v2 / Reinplace
|
|
0
|
69
|
September 9, 2025
|
|
How to understand the files generated by TORCH_COMPILE_DEBUG
|
|
0
|
57
|
September 4, 2025
|
|
Preserving bits like conj, neg during functionalization
|
|
0
|
36
|
June 12, 2025
|
|
Why adaptive_avg_pool1d became adaptive_avg_pool2d in make_fx?
|
|
1
|
105
|
April 21, 2025
|
|
How does torch.fx work with activation checkpointing?
|
|
2
|
835
|
February 13, 2025
|
|
Torch IR support for torch.exported graph
|
|
0
|
81
|
February 10, 2025
|
|
[torch.export] How to disable sym_size_int?
|
|
2
|
254
|
February 7, 2025
|
|
Aot_export_module captures more forward variables/parameters
|
|
2
|
105
|
January 23, 2025
|
|
Combine vmap, func.grad with tensordict
|
|
0
|
163
|
January 7, 2025
|
|
A node type in export IR graph
|
|
1
|
200
|
November 28, 2024
|
|
Any chance to preserve some ops while decomposing PT2E model?
|
|
1
|
89
|
November 16, 2024
|
|
Handling LSTM states as model's inputs/outputs using fx.symbolic_trace
|
|
0
|
59
|
November 5, 2024
|
|
Torch.fx.symbolic_trace with multiple GPUs
|
|
1
|
165
|
October 29, 2024
|
|
Is Export IR incompatible with fx.Interpreter?
|
|
1
|
219
|
October 8, 2024
|
|
How to turn a autograd.Function into custom_ops
|
|
0
|
61
|
August 24, 2024
|
|
‘len’ is not supported in symbolic tracing by default
|
|
7
|
2762
|
August 22, 2024
|
|
Add an input to a model with torch.compile
|
|
0
|
72
|
July 31, 2024
|
|
Noob here ! Need help with "RuntimeError: Input type (torch.FloatTensor)"
|
|
1
|
378
|
July 25, 2024
|
|
Tracking statistics of operators without overwriting forward pass?
|
|
0
|
103
|
July 22, 2024
|
|
Understanding torch.fx.traceback.preserve_node_meta()
|
|
1
|
359
|
July 10, 2024
|
|
When should I detect a fake tensor mode?
|
|
1
|
273
|
July 6, 2024
|
|
FX graph resorting a torch model's positional args based on execution order
|
|
1
|
173
|
June 11, 2024
|
|
Fx on split and cat combination
|
|
1
|
1203
|
June 10, 2024
|
|
Node Meta Data not propagated to replaced nodes while using fx subgraph rewrite?
|
|
2
|
625
|
June 5, 2024
|
|
StopIteration error
|
|
3
|
992
|
May 26, 2024
|
|
Can I use grad around a numpy code and compile it?
|
|
12
|
647
|
May 22, 2024
|