jrc2139/embeddinggemma-300m-qat-q8_0-unquantized-ONNX

ONNX Quantized versions of google/embeddinggemma-300m

This repository contains ONNX export and multiple quantized versions of google/embeddinggemma-300m-qat-q8_0-unquantized.

Usage

from sentence_transformers import SentenceTransformer

# Load Int8 model (ARM64 example)
model = SentenceTransformer(
    "jrc2139/embeddinggemma-300m",
    backend="onnx",
    model_kwargs={"file_name": "onnx/model_q4.onnx"},
    trust_remote_code=True
)