Enable mixtral 8x7b autotp #5257
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
|
Hi @mrwyattii @delock. Please kindly review. Thanks! |
25 tasks
@delock - do we want this merged in after your CPU autoTP PR? |
Hi @loadams , this can be merged before CPU autoTP workflow PR. I'll keep on working on that PR. |
mrwyattii
approved these changes
Mar 13, 2024
loadams
approved these changes
Mar 13, 2024
|
Hi @loadams. From the failure log it seems an env issue. Could you run the CI again to check if env issue? |
Hi @Yejing-Lai - yes we have a known env issue that we are working to resolve and will merge this PR when fixed. |
github-merge-queue
bot
removed this pull request from the merge queue due to failed status checks
rraminen
pushed a commit
to ROCm/DeepSpeed
that referenced
this pull request
May 9, 2024
This PR aims to enable mixtral 8x7b (MoE model) autotp. Co-authored-by: Logan Adams <[email protected]>
dbyoung18
pushed a commit
to dbyoung18/DeepSpeed
that referenced
this pull request
Jun 11, 2024
This PR aims to enable mixtral 8x7b (MoE model) autotp. Co-authored-by: Logan Adams <[email protected]>
loadams
added a commit
that referenced
this pull request
Dec 18, 2024
As title says. Default behavior of arctic model produces shape issues with AutoTP due to the MLP layer performing `w2 * act(w1*w3)`. However, method provided to fix Mixtral-7x8b in #5257 does not work since the MLP for Arctic is also used within a ModuleList for the MoE. This results in MLP weights hiding behind individual experts as layers `#.w#`, which is not caught by the fix in #5257. This adds the check directly within replace, where it can check for actual layer names for the `w2` key in the model to patch with `all_reduce`. --------- Signed-off-by: Daniel Huang <[email protected]> Co-authored-by: Olatunji Ruwase <[email protected]> Co-authored-by: Logan Adams <[email protected]>
siqi654321
pushed a commit
to siqi654321/DeepSpeed
that referenced
this pull request
Feb 7, 2025
As title says. Default behavior of arctic model produces shape issues with AutoTP due to the MLP layer performing `w2 * act(w1*w3)`. However, method provided to fix Mixtral-7x8b in deepspeedai#5257 does not work since the MLP for Arctic is also used within a ModuleList for the MoE. This results in MLP weights hiding behind individual experts as layers `#.w#`, which is not caught by the fix in deepspeedai#5257. This adds the check directly within replace, where it can check for actual layer names for the `w2` key in the model to patch with `all_reduce`. --------- Signed-off-by: Daniel Huang <[email protected]> Co-authored-by: Olatunji Ruwase <[email protected]> Co-authored-by: Logan Adams <[email protected]> Signed-off-by: siqi <[email protected]>
traincheck-team
pushed a commit
to traincheck-team/DeepSpeed
that referenced
this pull request
Feb 9, 2025
As title says. Default behavior of arctic model produces shape issues with AutoTP due to the MLP layer performing `w2 * act(w1*w3)`. However, method provided to fix Mixtral-7x8b in deepspeedai#5257 does not work since the MLP for Arctic is also used within a ModuleList for the MoE. This results in MLP weights hiding behind individual experts as layers `#.w#`, which is not caught by the fix in deepspeedai#5257. This adds the check directly within replace, where it can check for actual layer names for the `w2` key in the model to patch with `all_reduce`. --------- Signed-off-by: Daniel Huang <[email protected]> Co-authored-by: Olatunji Ruwase <[email protected]> Co-authored-by: Logan Adams <[email protected]>
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.
This PR aims to enable mixtral 8x7b (MoE model) autotp.