Pass context_fn to activation_checkpoint_params
#951
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
|
This pull request was exported from Phabricator. Differential Revision: D65360604 |
y-sq
changed the title
context_fn to activation_checkpoint_params
y-sq
added a commit
to y-sq/tnt
that referenced
this pull request
Dec 3, 2024
Summary: Add the policy "layer_based_auto_wrap_policy_float8_training". It skips the recompute of float8 scaling factor (a scaler) to improve the latency. To enable it, change the config file like: P1690229394 Reviewed By: yoyoyocmu Differential Revision: D65360604
y-sq
force-pushed
the
export-D65360604
branch
from
8d28ecc to
48aa7d7
Compare
|
This pull request was exported from Phabricator. Differential Revision: D65360604 |
Summary: Add the policy "layer_based_auto_wrap_policy_float8_training". It skips the recompute of float8 scaling factor (a scaler) to improve the latency. To enable it, change the config file like: P1690229394 Reviewed By: yoyoyocmu Differential Revision: D65360604
y-sq
force-pushed
the
export-D65360604
branch
from
48aa7d7 to
7c52c99
Compare
|
This pull request was exported from Phabricator. Differential Revision: D65360604 |
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Outdated suggestions cannot be applied.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:
We use
prepare_modulefromtorchtntin the model training, to wrap the model with activation checkpointing. We want to pass a customizedcontext_fn(to skip some ops from being recomputed) to the checkpointing wrapper, so added thecontext_fnfield inactivation_checkpoint_params.Test plan (run the training job e2e) is included in the original diff.
Reviewed By: yoyoyocmu
Differential Revision: D65360604