dist-keras
dist-keras copied to clipboard
cerndb
Distributed Deep Learning, with a focus on distributed training, using Keras and Apache Spark.
please I can run dis-kers on google Google colab
when run this code trainer = SingleTrainer(keras_model=model, worker_optimizer=optimizer, loss=loss, features_col="features_normalized", label_col="label", num_epoch=1, batch_size=32) trained_model = trainer.train(training_set) error display Py4JJavaError Traceback (most recent call last) in () 3 loss=loss, features_col="features_normalized", 4...
It is unclear how a graph plot can be generated showing the accuracy or loss over epochs. With keras, the history dictionary maintains the values across epochs. After drilling through...
I am trying to build LSTM model on time-series data. I am using MinMaxScaler to change range of features and target variable, then I reshaped data into 3d [samples, timestep,...
How to make results reproducible? I am using LSTM model to train time-series data, every time I run the code, I get different results. What wrong I am doing?
LSTM with DIST Keras : the problem is with preprocessing the sparkDF for the model : schema of the DF : each line have two matrix : FALSE and TRUE...
In the workflow example (https://github.com/cerndb/dist-keras/blob/master/examples/workflow.ipynb), getting error ```python --------------------------------------------------------------------------- TypeError Traceback (most recent call last) in () 1 trainer = DOWNPOUR(keras_model=model, worker_optimizer=optimizer, loss=loss, num_workers=num_workers, 2 batch_size=32, communication_window=5, learning_rate=0.05, num_epoch=1, ---->...
Hi @JoeriHermans, I am reading the documentation of this repository. My experimental setup runs on Theano for accuracy reproducibility. Now, I need to train my model in distributed fashion on...
Without this change, currently getting the error ```python --------------------------------------------------------------------------- AnalysisException Traceback (most recent call last) in () 1 # Only select the columns we need (less data shuffling) while training....
Addressing installation instructions not present here, but mentioned as needed here: https://github.com/cerndb/dist-keras/issues/36#issuecomment-335547831