Hello

As far as I can see the large model conatins filters of size 16, 24, 24, 40, 40, 40, 80, 80, 80, 80, 112, 112, 160, 160, 160. On the other hand, the official Keras MobileNetV2 implementation contains filters of size 16, 24, 24, 32, 32, 32, 64, 64, 64, 64, 96, 96, 96, 160, 160, 160, 320.

Why does this implementation here differ in terms of the number of filters?

Could you please advise on this error?

I tried to resolve it through "tf.compat.v1.", but unsure of this new output_shapes error ---->

line 34, in build_dataset
dataset["channels"] = ds_train.output_shapes["image"][-1].value

AttributeError: 'PrefetchDataset' object has no attribute 'output_shapes'

~Sanket

when I use paper parameters to train my task , the result is not good enough.

how to use NAS and NetAdapt to optimize mobilenet block and layer to adapt to other task?

hello,thanks for sharing your code.
But when I run the code, I find that the model is overfitting. For example, in cifar10 dataset, the train acc can reach 98%+, but val acc only 90%+.So what can I do?

Hi,
Thanks for sharing the code.
May I know what's the accuracy rate for MNIST and Cifar10 dataset ?

Dear @martinkersner ,

This is a question rather than an issue. I'd be grateful if you could give your thought.

I'm using your code to test a small (but old) project. The project was developed based on Queue-based and feed_dict mechanism of Session. That means I can only use your Model (build_mobilenetv3 function) without other Keras-based functions (model.compile, model.fit, model.evaluate, etc.)

The problem is that I need to set Keras's learning_phase to true during training, and false during evaluation or inference. I tried several ways like:

1. tf.keras.backend.set_learning_phase(True) / tf.keras.backend.set_learning_phase(False)
2. feed_dict = {tf.keras.backend.learning_phase() : 1} / feed_dict = {tf.keras.backend.learning_phase() : 0}
3. I also modified your code a bit such as:

model = BuildMobileNetV3(
        num_classes=5,
        width_multiplier=1.0,
        l2_reg=1e-5,
    )
input_shape=(150, 150, 3)
input_tensor = tf.keras.layers.Input(shape=input_shape)

And use,
model(input_tensor, training=True) / model(input_tensor, training=False)

However, none of the aforementioned methods works. The training cost and accuracy look good, but the validation accuracy's never been improved. I guess this is because of BN and dropout layers, but what else should I do beyond the above 3 approaches.

Thank you for your great work. I'm looking forward to discussing with you about the question.

Thanks.
Cuong

2022-03-27 16:28:09.999279: W tensorflow/core/framework/op_kernel.cc:1651] OP_REQUIRES failed at iterator_ops.cc:867 : Cancelled: Operation was cancelled
Traceback (most recent call last):
File "/home/hanwenxing/mobilenet_v3/train.py", line 113, in
main(args)
File "/home/hanwenxing/mobilenet_v3/train.py", line 81, in main
callbacks=callbacks,
File "/home/hanwenxing/anaconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py", line 727, in fit
use_multiprocessing=use_multiprocessing)
File "/home/hanwenxing/anaconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_arrays.py", line 675, in fit
steps_name='steps_per_epoch')
File "/home/hanwenxing/anaconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_arrays.py", line 300, in model_iteration
batch_outs = f(actual_inputs)
File "/home/hanwenxing/anaconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorflow_core/python/keras/backend.py", line 3476, in call
run_metadata=self.run_metadata)
File "/home/hanwenxing/anaconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1472, in call
run_metadata_ptr)
tensorflow.python.framework.errors_impl.InvalidArgumentError: 2 root error(s) found.
(0) Invalid argument: Expected dimension in the range [0, 0), but got -1
[[{{node metrics/acc/ArgMax}}]]
(1) Invalid argument: Expected dimension in the range [0, 0), but got -1
[[{{node metrics/acc/ArgMax}}]]
[[metrics/acc/Identity/_941]]
0 successful operations.
0 derived errors ignored.

Process finished with exit code 1

Can the program train its own models? I found that I couldn't finish it.

Hi! Thanks for providing the code! I wonder did you pretrain it on ImageNet and could you please share it?

Hello

Thanks for the code and the work. How can I remove the top layers (classification layer) so that I can add my own classification layers? Similar to when using MobileNetV2(weights=None, include_top=False) in tensorflow.keras.applications.

Firstly, thanks for the implementation of the MobileNetV3!

Do you have an interest in implementing the semantic segmentation head described in section 6.4? I am currently planning to implement that, so if you have an interest, we can work together to do it!

In layers.py line 41~56:

HardSigmoid() contains h-swish, and HardSwish() contains h-sigmoid.
Maybe simple mistake.

Regards.

when train cifar10, initial size is 9696 , and when downscale size is less than kernel 55 ,how to deal these channels?

there are only two datasets ?
how to train own datasets?

INFO 05-13 18:44:56 classify_flowers_mbv3.py:94 - Epoch: 0, iter: 0, loss: 1.6094379425048828, train_acc: 0.0
pred:  tf.Tensor([[[[nan nan nan nan nan]]]], shape=(1, 1, 1, 5), dtype=float32)
pred:  tf.Tensor([[[[nan nan nan nan nan]]]], shape=(1, 1, 1, 5), dtype=float32)
pred:  tf.Tensor([[[[nan nan nan nan nan]]]], shape=(1, 1, 1, 5), dtype=float32)
pred:  tf.Tensor([[[[nan nan nan nan nan]]]], shape=(1, 1, 1, 5), dtype=float32)
pred:  tf.Tensor([[[[nan nan nan nan nan]]]], shape=(1, 1, 1, 5), dtype=float32)
pred:  tf.Tensor([[[[nan nan nan nan nan]]]], shape=(1, 1, 1, 5), dtype=float32)

have you trained your implementation?

Hello

As part of the ModelCheckpoint callback I would like to save the mobilentv3 model (to hdf5) and later load the model. When doing so I'm getting a NotImplementedError. Is is not possible save the model?

Python 3.6.7
tensorflow-gpu 1.12.0