| .. |
|
__init__.py
|
[scan] better error handling and remove redundant tests (#137967)
|
2024-10-25 19:01:25 +00:00 |
|
associative_scan.py
|
Implementation of scan (#134102)
|
2024-09-10 04:51:16 +00:00 |
|
auto_functionalize.py
|
Generate slice.Tensor view operations instead of as_strided when split is used in the original program. (#137225)
|
2024-10-23 17:42:16 +00:00 |
|
cond.py
|
[cond] make cond not throw warnings on constant pred in eager mode (#138837)
|
2024-10-31 01:13:19 +00:00 |
|
effects.py
|
[effects] Turn off dtype promotion for with_effects lowering (#136039)
|
2024-09-16 16:14:05 +00:00 |
|
executorch_call_delegate.py
|
[hop] require hops to override __call__. (#134352)
|
2024-08-28 19:56:40 +00:00 |
|
flex_attention.py
|
[Flex Attention] Don't compute fill order to compute stride order just to get fill order back (#138376)
|
2024-10-22 18:40:39 +00:00 |
|
hints_wrap.py
|
[HOO] add hints_wrapper to support passing context hints (#132860)
|
2024-08-26 18:21:22 +00:00 |
|
invoke_subgraph.py
|
[invoke_subgraph] Change the joint_graph output signature to simplify min-cut partitioner (#139326)
|
2024-11-01 21:02:32 +00:00 |
|
map.py
|
[hop] preserve metadata in re-tracing hop subgraph by running with interpreter (#135159)
|
2024-09-05 21:36:56 +00:00 |
|
out_dtype.py
|
Make the __module__ name of HOO to be always "torch.ops.higher_order" (#132775)
|
2024-08-08 16:55:09 +00:00 |
|
run_const_graph.py
|
[hop] require hops to override __call__. (#134352)
|
2024-08-28 19:56:40 +00:00 |
|
scan.py
|
[scan] better error handling and remove redundant tests (#137967)
|
2024-10-25 19:01:25 +00:00 |
|
strict_mode.py
|
[Dynamo] Ensure torch function modes are dispatched on builtin ops (#137117)
|
2024-10-09 02:29:40 +00:00 |
|
torchbind.py
|
[hop] require hops to override __call__. (#134352)
|
2024-08-28 19:56:40 +00:00 |
|
triton_kernel_wrap.py
|
[Inductor] New Triton Attrs Descriptor Fixups (#138390)
|
2024-10-23 14:13:49 +00:00 |
|
utils.py
|
[hop] always trace subgraph with fake to support .item in eager mode (#138771)
|
2024-10-26 02:17:17 +00:00 |
|
while_loop.py
|
[Dynamo] Allow torch.cond() to handle emply arguments (#138190)
|
2024-10-26 15:26:21 +00:00 |
|
wrap.py
|
Allow fx graph caching higher order operators (opt-in) (#135877)
|
2024-09-24 17:23:09 +00:00 |
0e57f2b589