nkamiy/gemma3-4b-it-gguf

Gemma 3 4B IT — GGUF (Q5_K_M)

• Derived from google/gemma-3-4b-it. Modified: quantized to GGUF (Q5_K_M) using llama.cpp (commit fd62188).
• See NOTICE for license/usage terms.

Files

• gemma3-4b-it.Q5_K_M.gguf — text-only quantization
• gemma3-4b-it-mmproj.gguf — vision projector (optional, not quantized)
• Modelfile

How to use (Ollama - text-generation only)

ollama run hf.co/nkamiy/gemma3-4b-it-gguf:gemma3-4b-it.Q5_K_M.gguf

How to use (Ollama - image text to text)

1. Download the gguf files gemma3-4b-it.Q5_K_M.gguf and gemma3-4b-it-mmproj.gguf. And also download Modelfile. Put them in one folder
2. cd to the folder
3. Run the command as follows:

ollama create gemma3-4b-q5km -f Modelfile