circomlib-ml/models/reshape2d.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import Input, Reshape\n",
    "from tensorflow.keras import Model\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = Input(shape=(75,))\n",
    "x = Reshape((5, 5, 3))(inputs)\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, 75)]              0         \n",
      "                                                                 \n",
      " reshape (Reshape)           (None, 5, 5, 3)           0         \n",
      "                                                                 \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([[0.72327659, 0.97464849, 0.2592479 , 0.63799774, 0.89013732,\n",
       "        0.95867971, 0.06431743, 0.55685192, 0.77031965, 0.09982323,\n",
       "        0.10704737, 0.40713332, 0.57294341, 0.21883552, 0.22967276,\n",
       "        0.6221842 , 0.64159904, 0.684413  , 0.59126341, 0.88438877,\n",
       "        0.56715972, 0.93006015, 0.85704814, 0.79864291, 0.39604054,\n",
       "        0.30495253, 0.38333952, 0.69453548, 0.59207958, 0.30889659,\n",
       "        0.17302571, 0.41351124, 0.37527957, 0.43118255, 0.31526054,\n",
       "        0.12925303, 0.30129156, 0.73921834, 0.98336451, 0.03352392,\n",
       "        0.27839826, 0.6811155 , 0.96320785, 0.16882282, 0.68572833,\n",
       "        0.20924115, 0.30604142, 0.09080768, 0.63244108, 0.55914947,\n",
       "        0.60870048, 0.49377892, 0.9710362 , 0.12959508, 0.62162852,\n",
       "        0.26827067, 0.84771621, 0.40895646, 0.52476578, 0.48532215,\n",
       "        0.27144489, 0.19194784, 0.85410267, 0.11912042, 0.37034274,\n",
       "        0.25759208, 0.88306728, 0.98917787, 0.61814043, 0.49141046,\n",
       "        0.74162286, 0.81722887, 0.4728493 , 0.19955073, 0.42201694]])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = np.random.rand(1,75)\n",
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 44ms/step\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-04 00:14:24.910151: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[[[0.7232766 , 0.9746485 , 0.2592479 ],\n",
       "         [0.63799775, 0.8901373 , 0.9586797 ],\n",
       "         [0.06431743, 0.5568519 , 0.77031964],\n",
       "         [0.09982323, 0.10704737, 0.4071333 ],\n",
       "         [0.5729434 , 0.21883552, 0.22967276]],\n",
       "\n",
       "        [[0.6221842 , 0.64159906, 0.684413  ],\n",
       "         [0.5912634 , 0.88438874, 0.5671597 ],\n",
       "         [0.93006015, 0.85704815, 0.79864293],\n",
       "         [0.39604053, 0.30495253, 0.38333952],\n",
       "         [0.6945355 , 0.5920796 , 0.3088966 ]],\n",
       "\n",
       "        [[0.17302571, 0.41351125, 0.37527958],\n",
       "         [0.43118253, 0.31526053, 0.12925303],\n",
       "         [0.30129156, 0.73921835, 0.9833645 ],\n",
       "         [0.03352392, 0.27839825, 0.6811155 ],\n",
       "         [0.96320784, 0.16882282, 0.6857283 ]],\n",
       "\n",
       "        [[0.20924115, 0.30604142, 0.09080768],\n",
       "         [0.6324411 , 0.55914944, 0.60870045],\n",
       "         [0.4937789 , 0.9710362 , 0.12959507],\n",
       "         [0.6216285 , 0.26827067, 0.8477162 ],\n",
       "         [0.40895647, 0.5247658 , 0.48532215]],\n",
       "\n",
       "        [[0.2714449 , 0.19194785, 0.8541027 ],\n",
       "         [0.11912042, 0.37034273, 0.25759208],\n",
       "         [0.8830673 , 0.9891779 , 0.6181404 ],\n",
       "         [0.49141046, 0.74162287, 0.81722885],\n",
       "         [0.4728493 , 0.19955073, 0.42201695]]]], dtype=float32)"
      ]
     },
     "execution_count": 5,
     "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*1e36).round().astype(int).flatten().tolist(),\n",
    "    \"out\": (X*1e36).round().astype(int).flatten().tolist()\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"reshape2D_input.json\", \"w\") as f:\n",
    "    json.dump(in_json, f)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "keras2circom",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}