import matplotlib.pyplot as plt
from keras.callbacks import ModelCheckpoint
model_file_path = dir + f'{model_name}_'+'{epoch}.h5'
checkpoint = ModelCheckpoint(model_file_path, verbose=1, save_weights_only=False,period=10)
history = model.fit('图像数据','mask图数据', epochs = 100, batch_size= 4, validation_data = ('验证集图像数据', '验证集mask图数据', callbacks = [checkpoint]))
loss = history.history['loss']
df_temp = pd.DataFrame(loss, index=range(1,len(loss)+1), columns=['loss'])
df_temp.to_csv('记录loss曲线的数据的文件存储路径')
plt.plot(loss)
plt.title('Training loss')
plt.xlabel('epoch')
plt.ylabel('loss')
plt.savefig('loss曲线图片保存路径')
plt.show()
输出每个epoch的模型文件
from keras.callbacks import ModelCheckpoint
model_file_path = dir + f'{model_name}_'+'{epoch}.h5'
checkpoint = ModelCheckpoint(model_file_path, verbose=1, save_weights_only=False,period=10)
history = model.fit('图像数据','mask图数据', epochs = 100, batch_size= 4, validation_data = ('验证集图像数据', '验证集mask图数据', callbacks = [checkpoint]))
保存loss曲线信息
history = model.fit('图像数据','mask图数据', epochs = 100, batch_size= 4, validation_data = ('验证集图像数据', '验证集mask图数据', callbacks = [checkpoint]))
loss = history.history['loss']
df_temp = pd.DataFrame(loss, index=range(1,len(loss)+1), columns=['loss'])
df_temp.to_csv('记录loss曲线的数据的文件存储路径')