nunusadmqk/gemma-4-E4B-it-W8A8-INT8-v10-datafree

Gemma 4 E4B W8A8 INT8 v10 Datafree

This repository contains a W8A8 INT8 quantized conversion of google/gemma-4-E4B-it.

• —Base model: google/gemma-4-E4B-it
• —Quantization: W8A8 INT8
• —Format: compressed checkpoint / safetensors
• —Intended runtime: vLLM / Transformers-compatible loading where supported

This is an unofficial quantized conversion for deployment and testing.