How I can load a quantized model from my own host

#1
by yoeldcd - opened

I are trying to load smollm-360M-Instruct as 4q quantization, I specified dtype as '4q' on option object, but the pipeline show me an error of model smoll-360M-Instruct/onnx/model_merged_quantized.onnx not found

Screenshot_20240801-095137.png

I just configure my own host, but is not reading the correct quantized onnx file (model_q4.onnx)

Sign up or log in to comment