GSOC Log jiyeon's log

[Week 5] implement an estimator which uses simple regression (2)

Make a benchmarks In this week, I got the inference latency time for each model with a small dataset. (under 40 samples) As a result of checking roughly, the value of latency changed the most according to input shape, and the filter size also affected it a lot. And now I’m currently making larger dataset with below’s combinations. kernel_l... Read more

[Week 4] implement an estimator which uses simple regression

1. Seperate all the code and Test it on the device Successfully I can get the result what we want, the benchmark, latency time. But even though I seperate the code, but it still looks complicated. 😂 Now I’m getting the benchmarks of 19200 samples (still running). run_bazel.py import os import subprocess import sqlite3 import numpy as np impo... Read more

[Week 3] Collect initial benchmarks, start designing the database

Baseline Code import os import subprocess from itertools import product import numpy as np import pandas as pd import tensorflow as tf from tensorflow import keras def run_bazel(model_name): proc = subprocess.Popen( ['sh', 'gsoc_proj/run.sh', './gsoc_proj/MODELS/{}'.format(model_name)], stdout = subprocess.PIPE ) o... Read more

[Week 2] Have one-convolution models generated

How to get all Operations of a TFLite model In the interpreter.py code, especially _get_ops_details() method, it returns tensor index of each layer. Below is the code of getting all Conv2D layers inside tflite model and get the input/output tensor size. import tensorflow as tf SAVED_MODEL_PATH = "./densenet.tflite" interpreter = tf.lite.Int... Read more

[Week 1] Understand how the convolutions need to be generated

In this week, I tried to find out input/output size of each layers. Following is the example code from Ekaterina. const tflite::Interpreter* interpreter = ...; // Initialize interpreter for (int op_index : interpreter->execution_plan()) { const auto* op_and_reg = interpreter->node_and_registration(op_index); if (op_and_reg-&... Read more