AI Camera - IMX500 • Re: Step Quantization and Compression

What is the proper way to upload a model.pt file into python for quantization?
Also, what is wrong with this:

import torch
float_model_path = './yolo11n.pt'
float_model = torch.load(float_model_path)
#print(model)
float_model= float_model['model']
float_model.eval()

import numpy as np
num_calibration_batches = 10
def repr_datagen():
for _ in range(num_calibration_batches):
yield [np.random.random((4, 3, 640, 640))]

import model_compression_toolkit as mct
quantized_module, quantization_info = mct.ptq.pytorch_post_training_quantization(float_model, repr_datagen)

Statistics: Posted by sams1313 — Thu Dec 26, 2024 11:16 pm

---