Tensorflow raised graph execution error

Anyone familiar with tensorflow? 

Here's my code : 

```
try:
            model: object = tf.keras.Sequential([
                base_model,
                tf.keras.layers.Conv2D(32, 3, activation='relu'),
                tf.keras.layers.Dropout(0.2),
                tf.keras.layers.GlobalAveragePooling2D(),
                tf.keras.layers.Dense(36, activation='softmax')
            ])

            tf.compat.v1.enable_eager_execution()

            model.compile(
                optimizer=tf.keras.optimizers.Adam(
                    learning_rate=0.001,
                    beta_1=0.9,
                    beta_2=0.999,
                    epsilon=1e-07,
                    amsgrad=False,
                    name='Adam'
                ),
                loss='categorical_crossentropy',
                metrics=["accuracy"]
            )

            model.fit(
                train_generator,
                epochs=self.epochs,
                validation_data=value_generator,
                validation_steps=10,
                verbose=1
            )
        except Exception as e:
            print(str(e))
 ```

The error I got:

```
Graph execution error:

Detected at node 'categorical_crossentropy/softmax_cross_entropy_with_logits' defined at (most recent call last):
...
 return tf.nn.softmax_cross_entropy_with_logits(
Node: 'categorical_crossentropy/softmax_cross_entropy_with_logits'
logits and labels must be broadcastable: logits_size=[13,36] labels_size=[13,21]
         [[{{node categorical_crossentropy/softmax_cross_entropy_with_logits}}]] 

```
Has anyone encountered this before?