Tensorflow报错：InvalidArgumentError:indices[401,33] = 77571 is not in [0, 10000)

た入场券 2023-09-25 14:52 14阅读 0赞

在进行影评文本分类时，遇到一个报错，记录下解决过程。

#### 报错信息： ####

**InvalidArgumentError**：

indices\[401,33\] = 77571 is not in \[0, 10000)

\[\[\{ \{node embedding/embedding\_lookup\}\} = ResourceGather\[Tindices=DT\_INT32, dtype=DT\_FLOAT, validate\_indices=true, \_device="/job:localhost/replica:0/task:0/device:CPU:0"\](embedding/embeddings, embedding/Cast)\]\]

后在github上看了这边帖子[帖子][Link 1]，再分析了一下编写的RNN代码流程：

1、首先RNN网络第一层定义的的输入是10000

vocab_size = 10000
     
    model = tf.keras.Sequential()
    model.add(tf.keras.layers.Embedding(vocab_size, 16))
    model.add(tf.keras.layers.GlobalAveragePooling1D())
    model.add(tf.keras.layers.Dense(16, activation=tf.nn.relu))
    model.add(tf.keras.layers.Dense(1, activation=tf.nn.sigmoid))
     
    model.summary()

2、后续进行模型训练的时候，传入的数据为15000 \* 256长度的

# train_data.shape = (25000, 256)
    # train_labels.shape = (25000,)
     
    x_val = train_data[:10000]
    partial_x_train = train_data[10000:]
     
    y_val = train_labels[:10000]
    partial_y_train = train_labels[10000:]
     
    history = model.fit(partial_x_train,
                        partial_y_train,
                        epochs=40,
                        batch_size=512,
                        validation_data=(x_val, y_val),
                        verbose=1)

3、实际输入的长度与网络定义的长度不符，导致该错误发生

[Tensorflow报错：InvalidArgumentError:indices\[401,33\] = 77571 is not in \[0, 10000)\_Jayden Huang的博客-CSDN博客\_invalidargumenterror:][Tensorflow_InvalidArgumentError_indices_401_33_ _ 77571 is not in _0_ 10000_Jayden Huang_-CSDN_invalidargumenterror]

[Link 1]: https://github.com/tensorflow/tensorflow/issues/2734
[Tensorflow_InvalidArgumentError_indices_401_33_ _ 77571 is not in _0_ 10000_Jayden Huang_-CSDN_invalidargumenterror]: https://blog.csdn.net/u011585024/article/details/88938194