IBM Granite MoE Architecture #9438
Open
+92
−21
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
gabe-l-hart
force-pushed
the
GraniteMoE
branch
2 times, most recently
from
5f37be3
to
3219f58
Compare
2 tasks
compilade
reviewed
Sep 14, 2024
gabe-l-hart
force-pushed
the
GraniteMoE
branch
2 times, most recently
from
cd41666
to
1b235d0
Compare
This includes the addition of new tensor names for the new moe layers. These may not be correct at this point due to the need for the hack in gguf_writer.py to double-check the length of the shape for these layers. Branch: GraniteMoE Signed-off-by: Gabe Goodhart <[email protected]>
GraniteMoe has the same configuration deltas as Granite Branch: GraniteMoE Signed-off-by: Gabe Goodhart <[email protected]>
… and up After a lot of staring and squinting, it's clear that the standard mixtral expert implementation is equivalent to the vectorized parallel experts in granite. The difference is that in granite, the w1 and w3 are concatenated into a single tensor "input_linear." Rather than reimplementing all of the math on the llama.cpp side, the much simpler route is to just split this tensor during conversion and follow the standard mixtral route. Branch: GraniteMoE Co-Authored-By: [email protected] Signed-off-by: Gabe Goodhart <[email protected]>
GraniteMoE follows the mixtral architecture (once the input_linear layers are split into gate_exps/up_exps). The main delta is the addition of the same four multipliers used in Granite. Branch: GraniteMoE Signed-off-by: Gabe Goodhart <[email protected]>
gabe-l-hart
force-pushed
the
GraniteMoE
branch
from
1b235d0
to
2615459
Compare
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Outdated suggestions cannot be applied.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.
Dependencies
GraniteLMPR (IBM Granite Architecture #9412)Description
This PR introduces the
granitemoemodel architecture from IBM. It emulates thetransformerschanges in this PR.The
granitemoearchitecture follows a very similar pattern to thegranitearchitecture and its changes relative tollama. For the MoE variant, the base architecture ismixtral(MoE branch ofllamahere inllama.cpp). The same four additional multipliers are added (embeddings_multiplier,attention_multiplier,residual_multiplier, andlogits_scale).Testing
This PR can be tested using ibm/PowerMoE-3b from huggingface following the same testing steps used for
granite(here).