gemma-3-1b-it-gguf (fp32) base model from google original safetensors here tested on gguf-connector with nightly llama-cpp-python
