| from optimum.onnxruntime import ORTModelForTokenClassification, ORTQuantizer | |
| from optimum.onnxruntime.configuration import AutoQuantizationConfig | |
| onnx_model = ORTModelForTokenClassification.from_pretrained( | |
| "nickprock/bert-italian-finetuned-ner", | |
| export=True | |
| ) | |
| quantizer = ORTQuantizer.from_pretrained(onnx_model) | |
| dqconfig = AutoQuantizationConfig.avx512_vnni( | |
| is_static=False, | |
| per_channel=False | |
| ) | |
| model_quantized_path = quantizer.quantize( | |
| save_dir="bert-italian-ner-onnx-quantized-avx512", | |
| quantization_config=dqconfig, | |
| ) |