return none when use torch.load to load a quantize_model

[ev-day]

Describe the bug
while I use quantize_model.py(in script folder) to quantize ChatGLM3 model, it can be saved successfully, but when I try to load the model by torch.load(), it was return None.

To Reproduce
Steps to reproduce the behavior:
1.run 'python quantize_model.py -m .\chatglm3-6b -b 4 -o .\chatglm3-6b' to quantize the model, it will export below files

2. try to load model

3. return none

Except:
It should return a model object

[ev-day]

By another way, just I can quantize the model use 'model = intel_npu_acceleration_library.compile(model, dtype=int4) ', however, it will need much time to quantize and compile.