The tensorflow_2.0_tutorial from apm5

should word_num be len(word_index)+1?

Hi:

Thank you for providing such a good material.

In this file simple_example.py
the parameter word_num is 10000, while i run this code in my computer, an exception is raised:
InvalidArgumentError: indices[26,0] = 12089 is not in [0, 10000).

I think word_num should be set to len(word_index)+1, so i changed and tested again, and it worked.

I wonder if the change is right? or is there something i forget?

Thank you.

TypeError: Value passed to parameter 'input' has DataType uint8 not in list of allowed values: float16, bfloat16, float32, float64, int32

When running the ResNet_CIFAR.py, it will raise an error, after first epoch training.

100%|████████████████████████████████████████████████████████████████████████████████████████████████████████| 390/390 [00:42<00:00,  9.27it/s]
ce_loss:1.753367, l2_loss:0.201427, accuracy:0.355829
  0%|                                                                                                                   | 0/50 [00:00<?, ?it/s]
Traceback (most recent call last):
  File "tf2_train_cifar.py", line 190, in <module>
    test(model, test_images, test_labels)
  File "tf2_train_cifar.py", line 154, in test
    loss, prediction = test_step(model, x, y)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/def_function.py", line 568, in __call__
    result = self._call(*args, **kwds)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/def_function.py", line 615, in _call
    self._initialize(args, kwds, add_initializers_to=initializers)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/def_function.py", line 497, in _initialize
    *args, **kwds))
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/function.py", line 2389, in _get_concrete_function_internal_garbage_collected
    graph_function, _, _ = self._maybe_define_function(args, kwargs)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/function.py", line 2703, in _maybe_define_function
    graph_function = self._create_graph_function(args, kwargs)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/function.py", line 2593, in _create_graph_function
    capture_by_value=self._capture_by_value),
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/func_graph.py", line 978, in func_graph_from_py_func
    func_outputs = python_func(*func_args, **func_kwargs)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/def_function.py", line 439, in wrapped_fn
    return weak_wrapped_fn().__wrapped__(*args, **kwds)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/func_graph.py", line 968, in wrapper
    raise e.ag_error_metadata.to_exception(e)
TypeError: in converted code:

    tf2_train_cifar.py:116 test_step  *
        prediction = model(x, training=False)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/keras/engine/base_layer.py:778 __call__
        outputs = call_fn(cast_inputs, *args, **kwargs)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/keras/engine/network.py:717 call
        convert_kwargs_to_constants=base_layer_utils.call_context().saving)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/keras/engine/network.py:891 _run_internal_graph
        output_tensors = layer(computed_tensors, **kwargs)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/keras/engine/base_layer.py:778 __call__
        outputs = call_fn(cast_inputs, *args, **kwargs)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/keras/layers/convolutional.py:213 call
        outputs = self._convolution_op(inputs, self.kernel)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/nn_ops.py:1176 __call__
        return self.conv_op(inp, filter)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/nn_ops.py:662 __call__
        return self.call(inp, filter)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/nn_ops.py:252 __call__
        name=self.name)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/nn_ops.py:2052 conv2d
        name=name)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/gen_nn_ops.py:969 conv2d
        data_format=data_format, dilations=dilations, name=name)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/op_def_library.py:576 _apply_op_helper
        param_name=input_name)
    /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/op_def_library.py:61 _SatisfiesTypeConstraint
        ", ".join(dtypes.as_dtype(x).name for x in allowed_list)))

    TypeError: Value passed to parameter 'input' has DataType uint8 not in list of allowed values: float16, bfloat16, float32, float64, int32

Then I found that the error occurred in test function

tensorflow_2.0_tutorial/CNN/ResNet_CIFAR.py

Line 181 in 04bcfc3

x = images[i * test_batch_size: (i + 1) * test_batch_size, :, :, :]

Where after reading the images, unlike train function, after images_augment, the dtype will be tf.float32.
But for test function, it is still uint8 type.

The solution is to add a line:

x = tf.cast(x, tf.float32)

is_training in BatchNormalization

Hi,

I found the batch normalization in CNN/ResNet_CIFAR.py did not call the is_training to differentiate the training and test process. Is that correct? Thanks!

apm5 / tensorflow_2.0_tutorial Goto Github PK

tensorflow_2.0_tutorial's People

Contributors

Stargazers

Watchers

Forkers

tensorflow_2.0_tutorial's Issues

should word_num be len(word_index)+1?

TypeError: Value passed to parameter 'input' has DataType uint8 not in list of allowed values: float16, bfloat16, float32, float64, int32

is_training in BatchNormalization

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent