nullzero-live/gem3COMPILAR

1---2tags:3- gguf4- llama.cpp5- unsloth6- vision-language-model7base_model:8- google/gemma-3-4b-it9---10 11# gem3COMPILAR : GGUF12 13This model was finetuned and converted to GGUF format using [Unsloth](https://github.com/unslothai/unsloth).14 15**Example usage**:16- For text only LLMs:    `./llama.cpp/llama-cli -hf nullzero-live/gem3COMPILAR --jinja`17- For multimodal models: `./llama.cpp/llama-mtmd-cli -hf nullzero-live/gem3COMPILAR --jinja`18 19## Available Model files:20- `gemma-3-4b-it.Q8_0.gguf`21- `gemma-3-4b-it.F16-mmproj.gguf`22 23## ⚠️ Ollama Note for Vision Models24**Important:** Ollama currently does not support separate mmproj files for vision models.25 26To create an Ollama model from this vision model:271. Place the `Modelfile` in the same directory as the finetuned bf16 merged model283. Run: `ollama create model_name -f ./Modelfile`29   (Replace `model_name` with your desired name)30 31This will create a unified bf16 model that Ollama can use.32This was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth)33[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)