为 TFliteconverter 创建代表性数据集的正确方法是什么?

What is the correct way to create representative dataset for TFliteconverter?

我正在尝试使用 INT8 权重和激活来推断 tinyYOLO-V2。我可以使用 TFliteConverter 将权重转换为 INT8。对于 INT8 激活,我必须给出有代表性的数据集来估计比例因子。我创建此类数据集的方法似乎是错误的。

正确的步骤是什么?

def rep_data_gen():
    a = []
    for i in range(160):
        inst = anns[i]
        file_name = inst['filename']
        img = cv2.imread(img_dir + file_name)
        img = cv2.resize(img, (NORM_H, NORM_W))
        img = img / 255.0
        img = img.astype('float32')
        a.append(img)
    a = np.array(a)
    print(a.shape) # a is np array of 160 3D images
    img = tf.data.Dataset.from_tensor_slices(a).batch(1)
    for i in img.take(BATCH_SIZE):
        print(i)
        yield [i]
# https://www.tensorflow.org/lite/performance/post_training_quantization
converter = tf.lite.TFLiteConverter.from_keras_model_file("./yolo.h5")
converter.optimizations = [tf.lite.Optimize.DEFAULT]
converter.target_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8]
converter.inference_input_type = [tf.int8]
converter.inference_output_type = [tf.int8]
converter.representative_dataset=rep_data_gen
tflite_quant_model = converter.convert()

ValueError: Cannot set tensor: Got tensor of type STRING but expected type FLOAT32 for input 27, name: input_1

我使用你的代码读取数据集,发现错误:

img = img.astype('float32') 应该是

img = img.astype(np.float32)

希望对您有所帮助