Python: TensorBoard

Scritto ed eseguito sul portatile con Windows 10 – Effetto South Working

Utilizzo l’environment conda py3_tf

~$ conda activate py3_tf

TensorBoard

Basato Breast cancer Wisconsin costruiamo i log per tenere traccia

import pandas as pd
import numpy as np

from datetime import datetime
import os

from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler

import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Activation,Dropout
from tensorflow.keras.callbacks import EarlyStopping,TensorBoard

# df
df = pd.read_csv('cancer_classification.csv')

# X e y
X = df.drop('benign_0__mal_1',axis=1).values
y = df['benign_0__mal_1'].values

# train e test
X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.25,random_state=101)

# scaling
scaler = MinMaxScaler()
scaler.fit(X_train)
X_train = scaler.transform(X_train)
X_test = scaler.transform(X_test)

# early stopping
early_stop = EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=25)

# working directory
os.getcwd()

'F:\\Python\\Course 001'

# time
datetime.now().strftime("%Y-%m-%d--%H%M")

'2020-08-01--1701'

# definizione tensorboard
# WINDOWS: Use "logs\\fit"
# MACOS/LINUX: Use "logs\fit"

log_directory = 'logs\\fit'

# OPTIONAL: ADD A TIMESTAMP FOR UNIQUE FOLDER
timestamp = datetime.now().strftime("%Y-%m-%d--%H%M")
log_directory = log_directory + '\\' + timestamp

board = TensorBoard(
    log_dir=log_directory,
    histogram_freq=1, # dopo ciascuna epoch calcola i pesi
    write_graph=True,
    write_images=True,
    update_freq='epoch',
    profile_batch=2,
    embeddings_freq=1)

# definizione modello
model = Sequential()
model.add(Dense(units=30,activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(units=15,activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(units=1,activation='sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam')

# training modello (lanciarlo solo una volta nella stessa callback log_dir)
model.fit(x=X_train, 
          y=y_train, 
          epochs=600,
          validation_data=(X_test, y_test), verbose=0,
          callbacks=[early_stop,board]
          )

Epoch 00138: early stopping

<tensorflow.python.keras.callbacks.History at 0x2a852471d08>

Monitoring

Attivo tensorboard dal terminale

tensorboard --logdir logs\fit

Con il servizio attivo posso visitare http://localhost:6006/
Questo servizio di controllo e diagnostica è più utile per le convolutional neural networks.

print(log_directory)

logs\fit\2020-08-01--1701