17 lines
965 B
Plaintext
17 lines
965 B
Plaintext
The Livermore Big Artificial Neural Network toolkit (LBANN) is an open-source,
|
|
HPC-centric, deep learning training framework that is optimized to compose
|
|
multiple levels of parallelism.
|
|
|
|
LBANN provides model-parallel acceleration through domain decomposition to
|
|
optimize for strong scaling of network training. It also allows for composition
|
|
of model-parallelism with both data parallelism and ensemble training methods
|
|
for training large neural networks with massive amounts of data. LBANN is able
|
|
to advantage of tightly-coupled accelerators, low-latency high-bandwidth
|
|
networking, and high-bandwidth parallel file systems.
|
|
|
|
LBANN supports state-of-the-art training algorithms such as unsupervised,
|
|
self-supervised, and adversarial (GAN) training methods in addition to
|
|
traditional supervised learning. It also supports recurrent neural networks via
|
|
back propagation through time (BPTT) training, transfer learning, and
|
|
multi-model and ensemble training methods.
|