gemma4-e4b-openclaw-agent-gguf

gemma4-e4b-openclaw-agent-gguf

This repository contains the merged GGUF version of the model, optimized for efficient inference on CPU and GPU using llama.cpp.

Model Description

This is a GGUF format model specifically designed to run efficiently via llama-cpp-python and other compatible loaders. It contains the merged weights for local, low-resource deployment.

Usage with llama-cpp-python

from llama_cpp import Llama

# Load the model
llm = Llama(
    model_path="merged_model.gguf",
    n_ctx=2048, # Context window
    n_gpu_layers=0 # Increase this to offload layers to GPU
)

# Generate completion
output = llm(
    prompt="### Human: Hello!\n### Assistant:",
    max_tokens=256,
    stop=["### Human:"],
    temperature=0.7
)
print(output["choices"][0]["text"])