fyp/views/py/python_model_template.py

107 lines
2.6 KiB
Python
Raw Normal View History

import tensorflow as tf
import random
2023-10-03 11:55:22 +01:00
import pandas as pd
from tensorflow import keras
2023-10-02 21:15:31 +01:00
from tensorflow.data import AUTOTUNE
from keras import layers, losses, optimizers
2023-10-02 21:15:31 +01:00
DATA_DIR = "{{ .DataDir }}"
image_size = ({{ .Size }})
2023-10-03 11:55:22 +01:00
df = pd.read_csv("{{ .RunPath }}/train.csv", dtype=str)
keys = tf.constant(df['Id'].dropna())
values = tf.constant(list(map(int, df['Index'].dropna())))
table = tf.lookup.StaticHashTable(
initializer=tf.lookup.KeyValueTensorInitializer(
keys=keys,
values=values,
),
default_value=tf.constant(-1),
name="Indexes"
)
2023-10-02 21:15:31 +01:00
2023-10-03 11:55:22 +01:00
DATA_DIR_PREPARE = DATA_DIR + "/"
#based on https://www.tensorflow.org/tutorials/load_data/images
2023-10-02 21:15:31 +01:00
def pathToLabel(path):
2023-10-03 11:55:22 +01:00
path = tf.strings.regex_replace(path, DATA_DIR_PREPARE, "")
2023-10-02 21:15:31 +01:00
path = tf.strings.regex_replace(path, ".jpg", "")
2023-10-03 11:55:22 +01:00
path = tf.strings.regex_replace(path, ".png", "")
return table.lookup(tf.strings.as_string([path]))
#return tf.strings.as_string([path])
2023-10-02 21:15:31 +01:00
def decode_image(img):
2023-10-03 19:02:02 +01:00
img = tf.io.decode_png(img, channels={{.ColorMode}})
2023-10-02 21:15:31 +01:00
return tf.image.resize(img, image_size)
def process_path(path):
label = pathToLabel(path)
img = tf.io.read_file(path)
img = decode_image(img)
return img, label
def configure_for_performance(ds: tf.data.Dataset) -> tf.data.Dataset:
#ds = ds.cache()
ds = ds.shuffle(buffer_size= 1000)
ds = ds.batch(batch_size)
ds = ds.prefetch(AUTOTUNE)
return ds
def prepare_dataset(ds: tf.data.Dataset) -> tf.data.Dataset:
ds = ds.map(process_path, num_parallel_calls=AUTOTUNE)
ds = configure_for_performance(ds)
return ds
seed = random.randint(0, 100000000)
batch_size = 100
2023-10-02 21:15:31 +01:00
# Read all the files from the direcotry
list_ds = tf.data.Dataset.list_files(str(f'{DATA_DIR}/*'), shuffle=False)
image_count = len(list_ds)
list_ds = list_ds.shuffle(image_count, seed=seed)
val_size = int(image_count * 0.3)
train_ds = list_ds.skip(val_size)
val_ds = list_ds.take(val_size)
dataset = prepare_dataset(train_ds)
dataset_validation = prepare_dataset(val_ds)
model = keras.Sequential([
{{- range .Layers }}
{{- if eq .LayerType 1}}
layers.Rescaling(1./255),
{{- else if eq .LayerType 2 }}
2023-09-27 13:55:29 +01:00
layers.Dense({{ .Shape }}, activation="sigmoid"),
{{- else if eq .LayerType 3}}
layers.Flatten(),
{{- else }}
ERROR
{{- end }}
{{- end }}
])
2023-09-27 13:55:29 +01:00
model.compile(
2023-10-03 11:55:22 +01:00
loss=losses.SparseCategoricalCrossentropy(),
2023-09-27 13:55:29 +01:00
optimizer=tf.keras.optimizers.Adam(),
metrics=['accuracy'])
2023-09-29 13:27:43 +01:00
his = model.fit(dataset, validation_data= dataset_validation, epochs=50)
2023-09-27 13:55:29 +01:00
acc = his.history["accuracy"]
f = open("accuracy.val", "w")
f.write(str(acc[-1]))
f.close()
2023-09-27 21:20:39 +01:00
tf.saved_model.save(model, "model")
# model.save("model.keras", save_format="tf")