Repository Issue Activity (beta)

#98 Showcase / question: a board-proven offline language runtime on ESP32-C3, and whether some lightweight language capability may eventually crystallize beyond dense models

#101 Error fixes

#102 Error fixes

#106 [Architecture Flaw] Bi-directional Logic Retrieval Failure (The Reversal Curse) & Entity Biodata Solution

#107 ModuleNotFoundError: No module named 'gemma' when importing Gemma3ForMultimodalLM

#90 Refactor: Improve model classes and remove redundant parameters in gemma/model.py

#100 Support for Gemma 4?

#86 Ideas for gemma "4"

#97 Support for T5Gemma2

#92 Availability and support for Gemma3 270M models

#88 Bug: scripts/run_multimodal.py hangs in non-interactive environments like Docker

#82 Where can one find gemma-3 quantized weights?

#81 get_model_config raises ValueError when using official variant 4b-it

#84 Native support for pre-computed image embeddings

#87 Can we support multi GPUs when doing inference?

#78 No pytorch versions of gemma 3 models on HF

#74 Question about Rotary Embedding Sequence in Model Code vs. Diagrams

#85 Pretraining?

#79 Docker image comes with outdated Python version for Gemma-3 support

#50 Unable to reproduce MATH resulst

#11 MPS (Apple Silicon) Support

#77 Failed to cancel access request

#75 Bug: GemmaDecodeLayer __init__ is not passed required world_size, rank in model_xla

#72 How to solve the 'RESOURCE_EXHAUSTED' error when loading 'gemma2_instruct_2b_en' (the script is from kaggle and runs on colab with TPU)?

#60 Gemma finetuning formatting