Add numerical test comparing BetterDecoder and fairseq decoder #79576
Conversation
🔗 Helpful links
✅ No Failures (0 Pending)As of commit b4d1fbc (more details on the Dr. CI page): Expand to see more💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
|
This pull request was exported from Phabricator. Differential Revision: D37157391 |
|
Corresponding test for #79438 |
|
This pull request was exported from Phabricator. Differential Revision: D37157391 |
erichan1
force-pushed
the
export-D37157391
branch
from
122fa2a
to
0032648
Compare
@zrphercule Can you also approve #79438? It's the PR this one is stacked on top of. |
|
This pull request was exported from Phabricator. Differential Revision: D37157391 |
erichan1
force-pushed
the
export-D37157391
branch
from
0032648
to
07563bf
Compare
|
@pytorchbot merge -g |
|
@pytorchbot successfully started a merge job. Check the current status here |
|
Merge failed due to This PR has internal changes and must be landed via Phabricator |
…ch#79576) Summary: Pull Request resolved: pytorch#79576 Includes both forced and incremental decoding. Add a new file test_transformers.py to put transformers tests in and move away from huge monolithic test_nn.py. A todo item is to move existing transformer tests from test_nn.py to test_transformers.py. Add a numerical test comparing torch.nn._transformer_decoder_layer_fwd and fairseq decoder. Both decoders use the weights of a common nn.TransformerEncoder. Test Plan: ``` buck build --show-output mode/opt -c fbcode.enable_gpu_sections=true -c fbcode.nvcc_arch=a100 mode/inplace //caffe2/test:transformers ./fbcode/buck-out/gen/caffe2/test/transformers#binary.par ``` Test runs and passes! Differential Revision: D37157391 fbshipit-source-id: 3b3f1c7fdac8269278982e0dcc2d32ff6b63547d
…ch#79576) Summary: Pull Request resolved: pytorch#79576 Includes both forced and incremental decoding. Add a new file test_transformers.py to put transformers tests in and move away from huge monolithic test_nn.py. A todo item is to move existing transformer tests from test_nn.py to test_transformers.py. Add a numerical test comparing torch.nn._transformer_decoder_layer_fwd and fairseq decoder. Both decoders use the weights of a common nn.TransformerEncoder. Test Plan: ``` buck build --show-output mode/opt -c fbcode.enable_gpu_sections=true -c fbcode.nvcc_arch=a100 mode/inplace //caffe2/test:transformers ./fbcode/buck-out/gen/caffe2/test/transformers#binary.par ``` Test runs and passes! Reviewed By: mikekgfb Differential Revision: D37157391 fbshipit-source-id: 18ff84ba22e8b92208d4af97f266df0752c80d54
erichan1
force-pushed
the
export-D37157391
branch
from
07563bf
to
b4d1fbc
Compare
|
This pull request was exported from Phabricator. Differential Revision: D37157391 |
|
@pytorchbot merge |
|
@pytorchbot successfully started a merge job. Check the current status here |
|
Merge failed due to This PR has internal changes and must be landed via Phabricator |
|
Closing because code in this PR already committed in #79796 |
Summary:
Add a new file test_transformers.py to put transformers tests in and move away from huge monolithic test_nn.py. A todo item is to move existing transformer tests from test_nn.py to test_transformers.py.
Add a numerical test comparing torch.nn._transformer_decoder_layer_fwd and fairseq decoder. Both decoders use the weights of a common nn.TransformerEncoder. Contains both forced decoding and incremental decoding
Stacked on top of #79438
Test Plan:
Test runs and passes!
Differential Revision: D37157391