TensorFlow2.0 Model.fit auc不等于通过Model.predict然后在同一开发集中使用skle

Question

tf2.0，我的模型在下面的代码中显示，它的auc在训练时大约为0.62

def load_dataset(csv_path, shuffle=True):
return tf.data.experimental.make_csv_dataset(
csv_path,
batch_size=256,  
shuffle=shuffle,  
label_name='label',  
na_value='?',
num_epochs=1,
ignore_errors=True)
train_data = load_dataset('../data_demo/train/copy1_5train_index.csv')
train_data = train_data.map(embedding_train)
dev_data = load_dataset('../data_demo/dev/dev_index.csv')
dev_data = dev_data.map(embedding_dev)
model = tf.keras.Sequential([
tf.keras.layers.BatchNormalization(),
tf.keras.layers.Dense(128, activation='relu'),
tf.keras.layers.BatchNormalization(),
tf.keras.layers.Dense(1, activation='sigmoid'),
])
class_weight = {0: 0.19, 1: 0.81}
model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=[tf.keras.metrics.BinaryAccuracy(), tf.keras.metrics.AUC()])
checkpoint_save_path = 'model/dc.ckpt'
if os.path.exists(checkpoint_save_path + '.index'):
print('--------------load the model-------------')
model.load_weights(checkpoint_save_path)
cp_callback = tf.keras.callbacks.ModelCheckpoint(filepath=checkpoint_save_path,
save_weights_only=True,
save_best_only=False)
model.fit(train_data, epochs=5, class_weight=class_weight, validation_data=dev_data, callbacks=cp_callback)
model.summary()

但是，加载模型，然后用sklearn计算auc是0.5，代码显示在下面

from sklearn.metrics import roc_auc_score
def build_predict_result():
def load_dataset(csv_path, shuffle=True):
return tf.data.experimental.make_csv_dataset(
csv_path,
batch_size=1024, 
shuffle=shuffle,  
label_name='label', 
na_value='?',
num_epochs=1,
ignore_errors=True)
def load_vec():
news_vec = np.load('../data_demo/dev/dev_news_vec.npy')
user_vec = np.load('../data_demo/dev/dev_user_vec.npy')
return tf.constant(news_vec, dtype=tf.float32), tf.constant(user_vec, dtype=tf.float32)
nvec, uvec = load_vec()
def embedding(x, y):
return tf.concat([tf.nn.embedding_lookup(uvec, x['uindex']), tf.nn.embedding_lookup(nvec, x['nindex'])],
axis=1), y
dev_data = load_dataset('../data_demo/dev/dev_index.csv')
dev_data = dev_data.map(embedding)
model = tf.keras.Sequential([
tf.keras.layers.BatchNormalization(),
tf.keras.layers.Dense(128, activation='relu'),
tf.keras.layers.BatchNormalization(),
tf.keras.layers.Dense(1, activation='sigmoid'),
])
model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=['accuracy', 'AUC'])
checkpoint_save_path = 'model/dc.ckpt'
if os.path.exists(checkpoint_save_path + '.index'):
print('--------------load the model-------------')
model.load_weights(checkpoint_save_path)
result = model.predict(dev_data)
print(result.shape)
dev_df = pd.read_csv('../data_demo/dev/dev_index.csv')
dev_df['predict'] = result
dev_df.to_csv('dev_index_predict.csv', index=False)
build_predict_result()
def evaluate():
"""
计算ndcg
:return:
"""
df = pd.read_csv('dev_index_predict.csv')
auc = roc_auc_score(df['label'], df['predict'])
print('auc: ' + str(auc))
evaluate()

我找不到衡量计算auc的tf2.0源代码。通过用[1，0，0，1]，[0.4，0.2，0.3，0.5]等相同数据测试tf2.0和sklearn AUC函数，结果相等。

Answer 1

感谢大家关注我的问题。我的问题解决了。函数tf.data.experimental.make_csv_dataset中参数"shuffle"的默认值为True。因此，预测结果与原始csv数据不一致。将"shuffle"更改为False以解决我的问题。

def load_dataset(csv_path, shuffle=False):
return tf.data.experimental.make_csv_dataset(
csv_path,
batch_size=1024, 
shuffle=shuffle,  
label_name='label', 
na_value='?',
num_epochs=1,
ignore_errors=True)