Test case for EBC key-order change #2388
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Summary:
# context
current error:
```
1) torchrec.fb.ir.tests.test_serializer.TestSerializer: test_deserialized_device_vle
1) RuntimeError: Node ir_dynamic_batch_emb_lookup_default referenced nonexistent value id_list_features__values! Run Graph.lint() to diagnose such issues
While executing %ir_dynamic_batch_emb_lookup_default : [num_users=1] = call_function[target=torch.ops.torchrec.ir_dynamic_batch_emb_lookup.default](args = ([%id_list_features__values, None, %id_list_features__lengths, None], %floordiv, [4, 5]), kwargs = {})
Original traceback:
File "/data/users/hhy/fbsource/buck-out/v2/gen/fbcode/009ebbab256a7e75/torchrec/fb/ir/tests/__test_serializer__/test_serializer#link-tree/torchrec/fb/ir/tests/test_serializer.py", line 142, in forward
return self.sparse_arch(id_list_features)
File "torchrec/fb/ir/tests/test_serializer.py", line 446, in test_deserialized_device_vle
output = deserialized_model(features_batch_3.to(device))
File "torch/nn/modules/module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "torch/nn/modules/module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
File "torch/export/unflatten.py", line 482, in forward
tree_out = torch.fx.Interpreter(self, graph=self.graph).run(
File "torch/fx/interpreter.py", line 146, in run
self.env[node] = self.run_node(node)
File "torch/fx/interpreter.py", line 200, in run_node
args, kwargs = self.fetch_args_kwargs_from_env(n)
File "torch/fx/interpreter.py", line 372, in fetch_args_kwargs_from_env
args = self.map_nodes_to_values(n.args, n)
File "torch/fx/interpreter.py", line 394, in map_nodes_to_values
return map_arg(args, load_arg)
File "torch/fx/node.py", line 760, in map_arg
return map_aggregate(a, lambda x: fn(x) if isinstance(x, Node) else x)
File "torch/fx/node.py", line 768, in map_aggregate
t = tuple(map_aggregate(elem, fn) for elem in a)
File "torch/fx/node.py", line 768, in <genexpr>
t = tuple(map_aggregate(elem, fn) for elem in a)
File "torch/fx/node.py", line 772, in map_aggregate
return immutable_list(map_aggregate(elem, fn) for elem in a)
File "torch/fx/node.py", line 772, in <genexpr>
return immutable_list(map_aggregate(elem, fn) for elem in a)
File "torch/fx/node.py", line 778, in map_aggregate
return fn(a)
File "torch/fx/node.py", line 760, in <lambda>
return map_aggregate(a, lambda x: fn(x) if isinstance(x, Node) else x)
File "torch/fx/interpreter.py", line 391, in load_arg
raise RuntimeError(f'Node {n} referenced nonexistent value {n_arg}! Run Graph.lint() '
```
Differential Revision: D59238744
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D62604419 |
Summary: Pull Request resolved: pytorch#2388 # context * [post](https://fb.workplace.com/groups/1028545332188949/permalink/1042204770823005/) * this test case mimics the EBC key-order change after sharding {F1864056306} # details * it's a very simple model: EBC ---> KTRegroupAsDict * we generate two EBCs: ebc1 and ebc2, such that the table orders are different: ``` ebc1 = EmbeddingBagCollection( tables=[tb1_config, tb2_config, tb3_config], is_weighted=False, ) ebc2 = EmbeddingBagCollection( tables=[tb1_config, tb3_config, tb2_config], is_weighted=False, ) ``` * we export the model with ebc1 and unflatten the model, and then swap with ebc2 (you can think this as the the sharding process resulting a shardedEBC), so that we can mimic the key-order change as shown in the above graph * the test checks the final results after KTRegroupAsDict are consistent with the original eager model Reviewed By: PaulZhang12 Differential Revision: D62604419
TroyGarden
force-pushed
the
export-D62604419
branch
from
c6d3d9e
to
91a004b
Compare
|
This pull request was exported from Phabricator. Differential Revision: D62604419 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
context
{F1864056306}
details
Reviewed By: PaulZhang12
Differential Revision: D62604419