Can't increase the num_workers?

[AdityaSoni19031997]

PS Very New Flair User :)

Can't increase the num_workers...

To Reproduce

document_embeddings = DocumentRNNEmbeddings(word_embeddings,
                                            hidden_size=512,
                                            rnn_layers = 1,
                                            bidirectional=True,
                                            dropout = 0.23,
                                            reproject_words=True,
                                            reproject_words_dimension=256,
                                           )
classifier = TextClassifier(document_embeddings, label_dictionary=corpus.make_label_dictionary(), multi_label=False)
trainer = ModelTrainer(classifier, corpus)
trainer.train('./stacked_embeddings/',num_workers=16, max_epochs=10, mini_batch_size= 32, 
              learning_rate=0.1, anneal_factor=0.5, train_with_dev=True)

Expected behavior I expected to allow the Trainer.train() to increase the num_workers as such.. reference

Traceback

2019-05-28 18:15:22,483 Evaluation method: MICRO_F1_SCORE
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-22-6d2706307a23> in <module>
      1 trainer = ModelTrainer(classifier, corpus)
      2 trainer.train('./stacked_embeddings/',num_workers=16, max_epochs=10, mini_batch_size= 32, 
----> 3               learning_rate=0.1, anneal_factor=0.5, train_with_dev=True)

~/.local/lib/python3.7/site-packages/flair/trainers/trainer.py in train(self, base_path, evaluation_metric, learning_rate, mini_batch_size, eval_mini_batch_size, max_epochs, anneal_factor, patience, anneal_against_train_loss, train_with_dev, monitor_train, embeddings_in_memory, checkpoint, save_final_model, anneal_with_restarts, test_mode, param_selection_mode, **kwargs)
     80             weight_extractor = WeightExtractor(base_path)
     81 
---> 82         optimizer = self.optimizer(self.model.parameters(), lr=learning_rate, **kwargs)
     83         if self.optimizer_state is not None:
     84             optimizer.load_state_dict(self.optimizer_state)

TypeError: __init__() got an unexpected keyword argument 'num_workers'

But it's defined here,

Enviornments

• OS [Linux]:
• Version [e.g. flair-0.4.1]:

[alanakbik]

Hello @AdityaSoni19031997 did you install from pip? If so, this feature is not yet available. The feature is only available currently if you work on the master branch, which you could install like this:

!pip install --upgrade git+https://github.com/zalandoresearch/flair.git

[AdityaSoni19031997]

Yep I installed from pip! Thanks for the info guys, will pull the master branch :)

On Wed, 29 May 2019, 01:01 Alan Akbik, notifications@github.com wrote:

Hello @AdityaSoni19031997 https://github.com/AdityaSoni19031997 did you install from pip? If so, this feature is not yet available. The feature is only available currently if you work on the master branch, which you could install like this:

!pip install --upgrade git+https://github.com/zalandoresearch/flair.git

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/zalandoresearch/flair/issues/762?email_source=notifications&email_token=AFNPJJUZKU2JOYFMOXA4I2DPXWB7ZA5CNFSM4HQF7OTKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODWNF7DI#issuecomment-496656269, or mute the thread https://github.com/notifications/unsubscribe-auth/AFNPJJQ4ELYBVI52JG7DJX3PXWB7ZANCNFSM4HQF7OTA .