petrpan26/aivivn_1

Our submission for Aivivn Contest 1.

aivivn_1's Issues

File main_hierarchical.py dính lỗi value error

Traceback (most recent call last):
File "/content/drive/My Drive/Aivivn_1/main_hierarchical.py", line 247, in
args.fix_embed, args.add_embed, args.aug, args.aug_min_len, args.print_model, args.model
File "/content/drive/My Drive/Aivivn_1/main_hierarchical.py", line 117, in train_model
use_additive_emb = use_additive_emb
File "/content/drive/My Drive/Aivivn_1/scripts/rnn.py", line 198, in HRNN
word_lstm = Bidirectional(CuDNNLSTM(128))(embed)
File "/usr/local/lib/python3.6/dist-packages/keras/layers/wrappers.py", line 428, in call
return super(Bidirectional, self).call(inputs, **kwargs)
File "/usr/local/lib/python3.6/dist-packages/keras/engine/base_layer.py", line 408, in call
self.assert_input_compatibility(inputs)
File "/usr/local/lib/python3.6/dist-packages/keras/engine/base_layer.py", line 305, in assert_input_compatibility
str(K.ndim(x)))
ValueError: Input 0 is incompatible with layer bidirectional_1: expected ndim=3, found ndim=4

Something is wrong with Keras save model

Reference:

• keras-team/keras#4875

Upgrade CuDNN trên Google Colab

Anh có thể cho em biết làm thế nào để upgrade từ CuDNN 7.3.1 lên 7.6.2 trên Google Colab được ko ạ. Em chạy thử thấy code compile bằng CuDNN 7.6.2, nếu ko cài lại thì không chạy được nhiều mô hình trong hai file main và main_stack ạ.

main_stack.py bị dính value error

Traceback (most recent call last):   File "/content/drive/My Drive/Aivivn_1/main_stack.py", line 132, in     models_list, args.embedding, int(args.max), args.mix   File "/content/drive/My Drive/Aivivn_1/main_stack.py", line 64, in stack     for model in models_list   File "/content/drive/My Drive/Aivivn_1/main_stack.py", line 64, in     for model in models_list   File "/content/drive/My Drive/Aivivn_1/scripts/rnn.py", line 133, in SARNNKeras     x = Embedding(input_dim = max_features, output_dim = embed_size, weights = [embeddingMatrix], trainable = trainable)(inp)   File "/usr/local/lib/python3.6/dist-packages/keras/engine/base_layer.py", line 430, in call     self.set_weights(self._initial_weights)   File "/usr/local/lib/python3.6/dist-packages/keras/engine/base_layer.py", line 1051, in set_weights     'provided weight shape ' + str(w.shape)) ValueError: Layer weight shape (12001, 400) not compatible with provided weight shape (12001, 300)

SARNNKeras: TypeError: object of type 'Tensor' has no len()

Khi em chạy thử model này thì bị dính lỗi này anh ạ, anh có biết giải quyết như thế nào ko ạ?

File main_stack_hier.py bị dính lỗi TypeError

Traceback (most recent call last):
File "/content/drive/My Drive/Colab Notebooks/Aivivn_1/main_stack_hier.py", line 191, in
int(args.max), args.mix)
File "/content/drive/My Drive/Colab Notebooks/Aivivn_1/main_stack_hier.py", line 104, in stack
for model in models_list
File "/content/drive/My Drive/Colab Notebooks/Aivivn_1/main_stack_hier.py", line 104, in
for model in models_list
File "/content/drive/My Drive/Colab Notebooks/Aivivn_1/scripts/rnn.py", line 144, in SARNNKeras
)(x)
File "/usr/local/lib/python3.6/dist-packages/keras/engine/base_layer.py", line 451, in call
output = self.call(inputs, **kwargs)
File "/usr/local/lib/python3.6/dist-packages/keras_self_attention/seq_self_attention.py", line 169, in call
e = self._call_additive_emission(inputs)
File "/usr/local/lib/python3.6/dist-packages/keras_self_attention/seq_self_attention.py", line 223, in _call_additive_emission
q = K.tile(K.expand_dims(q, 2), K.stack([1, 1, input_len, 1]))
File "/usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py", line 2430, in tile
if len(n) < len(shape): # Padding the axis
TypeError: object of type 'Tensor' has no len()

OSError: Unable to create file (unable to open file: name = './models/2019-05-28-08:51:38-version/models.hdf5', errno = 22, error message = 'Invalid argument', flags = 13, o_flags = 302)

I got this error: OSError: Unable to create file (unable to open file: name = './models/2019-05-28-08:51:38-version/models.hdf5', errno = 22, error message = 'Invalid argument', flags = 13, o_flags = 302)
when calling:
python -m main -m VDCNN -e ./embeddings/smallFasttext.vi.vec --max 40000 --mix --prob
Here is the full log:

Using TensorFlow backend.
Number of train data: (16087, 1)
Here1
Folder already created
Folder already created
HERE2
2019-05-28 10:11:19.429735: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
WARNING: Logging before flag parsing goes to stderr.
W0528 10:11:19.639334  5676 deprecation.py:506] From C:\Users\Alice\Anaconda3\lib\site-packages\tensorflow\python\training\moving_averages.py:210: calling Zeros.__init__ (from tensorflow.python.op
s.init_ops) with dtype is deprecated and will be removed in a future version.
Instructions for updating:
Call initializer instance with the dtype argument instead of passing it to the constructor
HERE3
HERE4
W0528 10:11:21.114854  5676 deprecation.py:323] From C:\Users\Alice\Anaconda3\lib\site-packages\tensorflow\python\ops\array_grad.py:425: to_int32 (from tensorflow.python.ops.math_ops) is deprecate
d and will be removed in a future version.
Instructions for updating:
Use tf.cast instead.
Train on 15282 samples, validate on 805 samples
Epoch 1/100
15282/15282 [==============================] - 160s 11ms/step - loss: 0.6130 - acc: 0.6693 - f1: 0.5369 - val_loss: 0.8933 - val_acc: 0.6087 - val_f1: 0.1392

Epoch 00001: val_f1 improved from -inf to 0.13925, saving model to ./models/2019-05-28-10:10:58-version/models.hdf5
Traceback (most recent call last):
  File "C:\Users\Alice\Anaconda3\lib\runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "C:\Users\Alice\Anaconda3\lib\runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "F:\SOCIAL_MONITORING\AIVIVN\Aivivn_1\main.py", line 234, in <module>
    args.use_sim_dict, args.print_model, args.model)
  File "F:\SOCIAL_MONITORING\AIVIVN\Aivivn_1\main.py", line 123, in train_model
    batch_size=batch_size
  File "C:\Users\Alice\Anaconda3\lib\site-packages\keras\engine\training.py", line 1705, in fit
    validation_steps=validation_steps)
  File "C:\Users\Alice\Anaconda3\lib\site-packages\keras\engine\training.py", line 1255, in _fit_loop
    callbacks.on_epoch_end(epoch, epoch_logs)
  File "C:\Users\Alice\Anaconda3\lib\site-packages\keras\callbacks.py", line 77, in on_epoch_end
    callback.on_epoch_end(epoch, logs)
  File "C:\Users\Alice\Anaconda3\lib\site-packages\keras\callbacks.py", line 447, in on_epoch_end
    self.model.save(filepath, overwrite=True)
  File "C:\Users\Alice\Anaconda3\lib\site-packages\keras\engine\topology.py", line 2580, in save
    save_model(self, filepath, overwrite, include_optimizer)
  File "C:\Users\Alice\Anaconda3\lib\site-packages\keras\models.py", line 106, in save_model
    with h5py.File(filepath, mode='w') as f:
  File "C:\Users\Alice\Anaconda3\lib\site-packages\h5py\_hl\files.py", line 312, in __init__
    fid = make_fid(name, mode, userblock_size, fapl, swmr=swmr)
  File "C:\Users\Alice\Anaconda3\lib\site-packages\h5py\_hl\files.py", line 148, in make_fid
    fid = h5f.create(name, h5f.ACC_TRUNC, fapl=fapl, fcpl=fcpl)
  File "h5py\_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py\_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "h5py\h5f.pyx", line 98, in h5py.h5f.create
OSError: Unable to create file (unable to open file: name = './models/2019-05-28-10:10:58-version/models.hdf5', errno = 22, error message = 'Invalid argument', flags = 13, o_flags = 302)

I did dig into the log and I think that the file wasnot saved successfully before being read. The problem with keras saving model I think

Can't find index files './embeddings/annoy.pkl'

Anh ơi anh có thể chia sẻ file ./embeddings/annoy.pkl' được không ạ.