Updating the Gemma3 verification script to support safetensors checkpoints and numerical verification for FunctionGemma, a variant of Gemma3 270m.#927
Open
copybara-service[bot] wants to merge 1 commit intomainfrom
Conversation
…oints and numerical verification for FunctionGemma, a variant of Gemma3 270m. Key changes: -- Safetensors Support: Implemented a custom_loader within verify_gemma3.py to handle .safetensors files. This loader includes logic to remap tensor keys (e.g., fusing separate q_proj, k_proj, v_proj into qkv_proj and renaming normalization layers) to match the expectations of the litert_torch model builder and gemma_pytorch. -- Gemma3 270m Support: Added a get_config_for_270m function to define the configuration for the 270m variant (18 layers, 640 embedding dim, 4 heads). -- Enabling verification of FunctionGemma and Gemma3 270m models against the original implementation. These changes allow for successful numerical verification of Gemma3-1b (in safetensors format), Gemma3-270m, and FunctionGemma checkpoints. PiperOrigin-RevId: 865393370
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Updating the Gemma3 verification script to support safetensors checkpoints and numerical verification for FunctionGemma, a variant of Gemma3 270m.
Key changes:
-- Safetensors Support: Implemented a custom_loader within verify_gemma3.py to handle .safetensors files. This loader includes logic to remap tensor keys (e.g., fusing separate q_proj, k_proj, v_proj into qkv_proj and renaming normalization layers) to match the expectations of the litert_torch model builder and gemma_pytorch.
-- Gemma3 270m Support: Added a get_config_for_270m function to define the configuration for the 270m variant (18 layers, 640 embedding dim, 4 heads).
-- Enabling verification of FunctionGemma and Gemma3 270m models against the original implementation.
These changes allow for successful numerical verification of Gemma3-1b (in safetensors format), Gemma3-270m, and FunctionGemma checkpoints.