Sequence Parallel Forward Transformer #3338
Merged
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
|
!build |
samnordmann
approved these changes
Nov 18, 2024
wujingyue
reviewed
Nov 18, 2024
wujingyue
reviewed
Nov 18, 2024
| @@ -1074,11 +1305,11 @@ TEST_P(DistributedTransformerTest, Forward) { | |||
| auto ln_input = castOp(DataType::Float, x); | |||
| auto ln0 = layer_norm(ln_input, norm_shape, ln0_w, ln0_b, eps); | |||
| auto mha_in = castOp(dtype, ln0.output); | |||
| auto mha_out = mha(mha_in, mha_w0, mha_b0, mha_w1, mha_b1, mesh)[3]; | |||
| auto mha_out = mha(mha_in, mha_w0, mha_b0, mha_w1, mha_b1, mesh).output; | |||
wujingyue
approved these changes
Nov 19, 2024
|
!build |
|
check |
|
!test |
Priya2698
pushed a commit
that referenced
this pull request
Nov 20, 2024
Sequence parallel forward transformer layer and multi-headed attention tests. 1. Cleans up sharding annotations in Forward fusion definitions. Only sharding changes and inputs are explicitly sharded. 2. Updates output of mha and mlp to be a struct with named TVs to make code more readable. 3. Dropout probability is temporarily set to 0. This will be fixed in a later PR to use philox seed and offset with validation.
jacobhinkle
pushed a commit
that referenced
this pull request
Dec 3, 2024
Sequence parallel forward transformer layer and multi-headed attention tests. 1. Cleans up sharding annotations in Forward fusion definitions. Only sharding changes and inputs are explicitly sharded. 2. Updates output of mha and mlp to be a struct with named TVs to make code more readable. 3. Dropout probability is temporarily set to 0. This will be fixed in a later PR to use philox seed and offset with validation.
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.
Sequence parallel forward transformer layer and multi-headed attention tests.