Subversion Repositories keras_to_fpga

#

import numpy as np

import tensorflow as tf

from tensorflow.keras.datasets import cifar10

import matplotlib.pyplot as plt

import matplotlib.gridspec as gridspec

import shutil

import struct

import os

# ---------------------------------------------------------------

def plot_history(history, name):

  # summarize history for accuracy

  plt.plot(history.history['sparse_top_k_categorical_accuracy'])

  plt.title('model accuracy')

  plt.ylabel('accuracy')

  plt.xlabel('epoch')

  plt.savefig(str(name) + '_model_accuracy.png')

  plt.close()

  # summarize history for loss

  plt.plot(history.history['loss'])

  plt.title('model loss')

  plt.ylabel('loss')

  plt.xlabel('epoch')

  plt.savefig(str(name) + '_model_loss.png')

  plt.close()

# ---------------------------------------------------------------

def show_the_image(image, title=None):

  # plt.imshow(image, interpolation='bilinear')

  plt.imshow(image, cmap=plt.cm.gray)

  if title is not None:

    plt.title(title)

  plt.show()

# ---------------------------------------------------------------

def save_the_image(image, filename, title=None):

  # plt.imshow(image, interpolation='bilinear')

  plt.imshow(image)

  if title is not None:

    plt.title(title)

  plt.savefig(filename)

  plt.close()

# -------------------------------------------------------

def grid_plot(im):

  nrow = im.shape[0]

  ncol = im.shape[1]

  fig = plt.figure(figsize=(ncol+1, nrow+1))

  gs = gridspec.GridSpec(nrow, ncol,

           wspace=0.02, hspace=0.02,

           top=1.-0.5/(nrow+1), bottom=0.5/(nrow+1),

           left=0.5/(ncol+1), right=1-0.5/(ncol+1))

  for i in range(nrow):

    for j in range(ncol):

      ax= plt.subplot(gs[i,j])

      ax.imshow(im[i][j], cmap=plt.get_cmap('gray'))

      plt.axis('off')

  plt.show()

# -------------------------------------------------------

def load_mnist():

  mnist = tf.keras.datasets.mnist  # mnist is a dataset of 28x28 images of handwritten digits and their labels

  # input image dimensions

  img_rows, img_cols = 28, 28

  # the data, split between train and test sets

  (x_train, y_train), (x_test, y_test) = mnist.load_data()

  x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)

  x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)

  x_train = x_train.astype('float32')

  x_test = x_test.astype('float32')

  x_train /= 255

  x_test /= 255

  print('x_train shape:', x_train.shape)

  print(x_train.shape[0], 'train samples')

  print(x_test.shape[0], 'test samples')

  return (x_train, y_train),(x_test, y_test)

# -------------------------------------------------------

def load_cifar10():

  # The data, split between train and test sets:

  (x_train, y_train), (x_test, y_test) = cifar10.load_data()

  # print('x_train shape:', x_train.shape)

  # print(x_train.shape[0], 'train samples')

  # print(x_test.shape[0], 'test samples')

  return (x_train, y_train),(x_test, y_test)

# -------------------------------------------------------

def float_to_hex(f):

    return format(struct.unpack('<I', struct.pack('<f', f))[0], 'x')

# -------------------------------------------------------

def save_float_to_raw(a, name, index=None):

  if index:

    file_name = name + '_' + str(index) + '.raw'

  else:

    file_name = name + '.raw'

  print('save_float_to_raw() |', file_name)

  a = a.astype('float32')

  with open(file_name, "bw") as fh:

    a.flatten().tofile(fh)

# -------------------------------------------------------

def write_conv2d_kernel(layer, dir='weights'):

  w = layer.get_weights()

  for i in range(layer.input_shape[3]):

    for o in range(layer.output_shape[3]):

      file_name = dir + '/' + layer.name + '-' + str(i)  + '_' + str(o) + '.txt'

      print(file_name)

      with open(file_name, "w") as text_file:

        for y in range(0, 3):

          for x in range(0, 3):

            print(float_to_hex(w[0][x,y,i,o]), file=text_file)

# -------------------------------------------------------

def write_dense_weights(layer, dir='weights'):

  w = layer.get_weights()

  for y in range(0, w[0].shape[1]):

    file_name = dir + '/' + layer.name + '_' + str(y) + '.txt'

    print(file_name)

    with open(file_name, "w") as text_file:

      for x in range(0, w[0].shape[0]):

        print(float_to_hex(w[0][x][y]), file=text_file)

      print(float_to_hex(w[1][y]), file=text_file)