SymIntify torchrec variable batch size path #2394
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
✅ Deploy Preview for pytorch-fbgemm-docs ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 5, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing. SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. Removing int() convertions for SymInt. Adding torch._check() for VB parameters Differential Revision: D54554735
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
d1848ef
to
2fb1f95
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
2fb1f95
to
e702c36
Compare
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 6, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing. SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. Removing int() convertions for SymInt. Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 6, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing. SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. Removing int() convertions for SymInt. Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
e702c36
to
f572fb8
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 6, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing. SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. Removing int() convertions for SymInt. Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
f572fb8
to
0cf36ec
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
0cf36ec
to
6ebe9f5
Compare
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 7, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing. SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. Removing int() convertions for SymInt. Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 7, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing. SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. Removing int() convertions for SymInt. Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
6ebe9f5
to
f7daeea
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 7, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing: 1/ SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. The main idea of `adjust_info_B_num_bits` is to fit BatchSize (B) and number of Tables (T) into 32 bit. Default value for B_bits == 26, T_bits == 6. `adjust_info_B_num_bits` will at runtime change those bit numbers to fit specified T and B using control flow and bit shifts to check that T and B fits in bits. We have flows with T ~ 600, so picking default values for dynamo T_bits == 10 and B_bits == 22, guarding to fail in runtime if those bit numbers does not fit specified T and B. For Dynamo we will use constant values for now. 2/ fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. 3/ Removing int() convertions for SymInt. 4/ Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
f7daeea
to
55c95c4
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
IvanKobzarev
pushed a commit
to IvanKobzarev/FBGEMM
that referenced
this pull request
Mar 7, 2024
Summary: Variable Batch parameters are SymInt in dynamo tracing: 1/ SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. The main idea of `adjust_info_B_num_bits` is to fit BatchSize (B) and number of Tables (T) into 32 bit. Default value for B_bits == 26, T_bits == 6. `adjust_info_B_num_bits` will at runtime change those bit numbers to fit specified T and B using control flow and bit shifts to check that T and B fits in bits. We have flows with T ~ 600, so picking default values for dynamo T_bits == 10 and B_bits == 22, guarding to fail in runtime if those bit numbers does not fit specified T and B. For Dynamo we will use constant values for now. 2/ fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. 3/ Removing int() convertions for SymInt. 4/ Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
55c95c4
to
f770083
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
Summary: Variable Batch parameters are SymInt in dynamo tracing: 1/ SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values. The main idea of `adjust_info_B_num_bits` is to fit BatchSize (B) and number of Tables (T) into 32 bit. Default value for B_bits == 26, T_bits == 6. `adjust_info_B_num_bits` will at runtime change those bit numbers to fit specified T and B using control flow and bit shifts to check that T and B fits in bits. We have flows with T ~ 600, so picking default values for dynamo T_bits == 10 and B_bits == 22, guarding to fail in runtime if those bit numbers does not fit specified T and B. For Dynamo we will use constant values for now. 2/ fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum. 3/ Removing int() convertions for SymInt. 4/ Adding torch._check() for VB parameters Reviewed By: ezyang Differential Revision: D54554735
IvanKobzarev
force-pushed
the
export-D54554735
branch
from
f770083
to
6dd5241
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54554735 |
|
This pull request has been merged in 9751a61. |
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:
Variable Batch parameters are SymInt in dynamo tracing.
SymInt does not support bit shifts => Skipping adjust_info_B_num_bits logic for dynamo case (when SymInt arrive into kernel) defaulting the values.
fbcode/deeplearning/fbgemm/fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py - changing List comprehension with python sum to torch.sum.
Removing int() convertions for SymInt.
Adding torch._check() for VB parameters
Differential Revision: D54554735