Bfloat16 version of the Granite3 dense 8B instruct model.

tools

9 2 months ago

Readme

Bfloat16 version of the Granite3 dense 8B instruct model. Requires Ollama >v0.5.5 and BF16 supporting GPU (native or BF16 to FP32 on the fly llama.cpp conversion)

Full model info at https://ollama.com/library/granite3.1-dense and https://huggingface.co/ibm-granite/granite-3.1-8b-instruct/