circomlib-ml/models/sumPooling2d.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import Input, AveragePooling2D, Lambda\n",
    "from tensorflow.keras import Model\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = Input(shape=(5,5,3))\n",
    "x = AveragePooling2D(pool_size=2)(inputs)\n",
    "x = Lambda(lambda x: x*4)(x)\n",
    "model = Model(inputs, x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 5, 5, 3)]         0         \n",
      "_________________________________________________________________\n",
      "average_pooling2d (AveragePo (None, 2, 2, 3)           0         \n",
      "_________________________________________________________________\n",
      "lambda (Lambda)              (None, 2, 2, 3)           0         \n",
      "=================================================================\n",
      "Total params: 0\n",
      "Trainable params: 0\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[[0.83128186, 0.15650764, 0.23798145],\n",
       "         [0.00277366, 0.8374127 , 0.95278315],\n",
       "         [0.3074389 , 0.21931738, 0.14886067],\n",
       "         [0.13590018, 0.98728255, 0.12085182],\n",
       "         [0.47212572, 0.51380922, 0.74891219]],\n",
       "\n",
       "        [[0.74680338, 0.2533205 , 0.5039968 ],\n",
       "         [0.14475403, 0.00791911, 0.4361197 ],\n",
       "         [0.69925568, 0.77507624, 0.40388991],\n",
       "         [0.29508251, 0.99375606, 0.84959701],\n",
       "         [0.88844918, 0.33910189, 0.9617212 ]],\n",
       "\n",
       "        [[0.76480625, 0.591287  , 0.0714191 ],\n",
       "         [0.94371681, 0.1695303 , 0.4476252 ],\n",
       "         [0.54372616, 0.83818804, 0.95211573],\n",
       "         [0.30485104, 0.15165265, 0.94709317],\n",
       "         [0.90827137, 0.58854675, 0.01857002]],\n",
       "\n",
       "        [[0.70123418, 0.43090173, 0.7096038 ],\n",
       "         [0.20637783, 0.20096581, 0.22956612],\n",
       "         [0.81978383, 0.16775403, 0.67412096],\n",
       "         [0.1011535 , 0.35596916, 0.36702071],\n",
       "         [0.5874605 , 0.79341372, 0.93292159]],\n",
       "\n",
       "        [[0.77997124, 0.46311399, 0.5465576 ],\n",
       "         [0.20406287, 0.37547625, 0.59862253],\n",
       "         [0.52933135, 0.84249092, 0.02969684],\n",
       "         [0.29114617, 0.10405779, 0.5359062 ],\n",
       "         [0.25197146, 0.83297465, 0.67025403]]]])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = np.random.rand(1,5,5,3)\n",
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[[1.7256129, 1.2551599, 2.1308813],\n",
       "         [1.4376774, 2.9754324, 1.5231993]],\n",
       "\n",
       "        [[2.6161351, 1.3926848, 1.4582142],\n",
       "         [1.7695144, 1.5135639, 2.9403505]]]], dtype=float32)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = model.predict(X)\n",
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "in_json = {\n",
    "    \"in\": (X*1000).round().astype(int).flatten().tolist()\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "out_json = {\n",
    "    \"out\": (y*1000).round().astype(int).flatten().tolist()\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"sumPooling2D_input.json\", \"w\") as f:\n",
    "    json.dump(in_json, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"sumPooling2D_output.json\", \"w\") as f:\n",
    "    json.dump(out_json, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tf24",
   "language": "python",
   "name": "tf24"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.6"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}