0.14.14

This release includes:

• Allow add_embedding_path to be called via stdout API
• Print traceback of failed stdout API to stderr per default

0.14.13

This release includes several bugfixes and new features

• Allow building Docker images with files from the job.
• Allow to set install_files configuration option to add files from the job to the Docker image, when image is defined.
• Fixed handling online check in job end routine.
• Changed git reset to git checkout to not modify git's HEAD, this
  led to lock issues by Git itself when 2 processes try to reset
  at the same time.

0.14.12

This bugfix releases includes

• Drastically improvements for Python3 and word2vec parsing: 11minutes vs 5seconds.
• Performance improvements in reading and parsing the stdout of commands, especially ones without newlines.

0.14.9

This release fixed a bug introduced in 0.14.8

docker: Error response from daemon: Unknown runtime specified nvidia

0.14.6

This release fixes an issue with new timed channels. Please update to see the correct last traces of each channel in the job list on the web UI.

0.14.3

GC and close correctly all ssh connections. This release fixes OSError: Too many open files and decreases the load on AETROS Trainer servers.

0.14.1

fixed running jobs in cluster from local directory

0.10.0

Please see http://aetros.com/blog/Releases/2-09-2017-update-to-2017-4-version

0.9.3

• Changed message protocol to MsgPack
• Made Keras 2 compatible
• Fully Python3 compatibility
• Fixed issue with generating image insights for Keras 1 in Tensorflow
• Added new file upload method
• Added new batch progress method
• Many bug fixes and improvements

See http://aetros.com/docu/python-sdk/api/aetros.backend.jobbackend for full API documentation

Changed Keras integration API

Please use now our "Keras Callback"

import aetros.backend
job = aetros.backend.start_job('name/model')

(X_train, y_train), (X_test, y_test) = mnist.load_data()

model = Sequential()
model.add(...)
model.compile(...)

# version simplest
aetros_callback = job.create_keras_callback(model)

# or version with insights images and confusion matrix
aetros_callback = job.create_keras_callback(
model,
    insights=True, insights_x=X_train[0],
    confusion_matrix=True, validation_data=(X_test, Y_test)
)
model.fit(
    X_train, Y_train, batch_size=16, nb_epoch=50,
    validation_data=(X_test, Y_test), callbacks=[aetros_callback]
)

0.7.0

0.7.0 release