File size: 624 Bytes
aabc480 5b3fa81 aabc480 6e52db4 1393c58 5b3fa81 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 | ---
license: mit
language:
- en
library_name: keras
tags:
- tests
- tf_model_zoo
---
# MobileNetV2
This model - MobileNetV2 is generated from `tf.keras.applications`
using [tf_model_generator.py](https://github.com/syna-astra-dev/iree-synaptics-synpu/blob/main/tests/model_generator/tf_model_generator.py).
The dataset for int8 quantization is done using random data.
Available formats: int16, int8, float32, float16
The `kaggle.tflite` model is the [original quantized model](https://www.kaggle.com/models/tensorflow/mobilenet-v2/tfLite/1-0-224-quantized)
published by Google. This model uses v1 tflite quantization.
|