IBM Granite Architecture#9412

Merged

ggerganov merged 9 commits intoggml-org:masterfrom

gabe-l-hart:GraniteLM

Sep 17, 2024

Commits on Sep 16, 2024

feat(gguf-py): Add Granite model and params to gguf-py
gabe-l-hart
committed
feat(convert_hf_to_gguf): Add registration and param setup for Granite
gabe-l-hart
committed
feat(llama.cpp): Add config parsing for Granite multiplier params
gabe-l-hart
committed
feat(llama.cpp): First pass at full port of granite deviations from llama
gabe-l-hart
committed
fix(llama.cpp): Determine granite language 3b instruct by vocab size
gabe-l-hart
committed
fix(convert_hf_to_gguf): Use LlamaModel as base for GraniteModel
gabe-l-hart
committed
fix(llama.cpp): Switch Granite param names to use _scale for consistency
gabe-l-hart
committed
fix(convert_hf_to_gguf/gguf-py): _multiplier -> _scale
gabe-l-hart
committed
fix(llama.cpp): Use separate switch clause for granite in llm_load_hparams
gabe-l-hart
committed