Difference between revisions of "Eosc/easybuild/benchmark"

From Nordic Language Processing Laboratory
Jump to: navigation, search
(NumPy)
(TensorFlow)
Line 32: Line 32:
 
We use [https://github.com/ltgoslo/simple_elmo_training training an ELMo model on 2 GPUs] as a benchmark.  
 
We use [https://github.com/ltgoslo/simple_elmo_training training an ELMo model on 2 GPUs] as a benchmark.  
 
Any plain text corpus can be fed to this code.
 
Any plain text corpus can be fed to this code.
 +
 +
== TF with OpenBLAS on Saga ==
 +
A corpus of 59 579 878 word tokens. Vocabulary size 10 000. 1 epoch.
 +
 +
Training time: '''01:49:15'''
 +
 +
== TF with OpenMKL on Saga ==
 +
A corpus of 59 579 878 word tokens. Vocabulary size 10 000. 1 epoch.
 +
 +
Training time: '''01:48:21'''

Revision as of 23:36, 22 November 2020

Background

In the context of the EOSC-Nordic EasyBuild pilot, the following page provides instructions for how to benchmark different software configurations on typical problems that are likely to affect NLPL users. Relevant variation will typically contrast pre-compiled binary installations (e.g. `pip` wheels) vs. locally compiled modules, where architecture-specific optimizations are enabled and optimized libraries (e.g. Intel MKL) are used.

NumPy

We use this Python script which runs multiple random matrix multiplications and singular value decompositions (SVD). Only CPU is employed.

OpenBLAS on Saga

$ module use -a /cluster/shared/nlpl/software/easybuild_ak/easybuild/install/modules/all/
$ module load numpy/1.18.1-foss-2019b-Python-3.7.4
$ python3 tests/numpy/numpy_test.py
Multiplication took 78 seconds.
SVD took 60 seconds.

IMKL on Saga

$ module use -a /cluster/shared/nlpl/software/easybuild_ak/easybuild/install/modules/all/
$ module load NLPL-numpy/1.18.1-gomkl-2019b-Python-3.7.4
$ python3 tests/numpy/numpy_test.py
Multiplication took 55 seconds.
SVD took 49 seconds.

TensorFlow

We use training an ELMo model on 2 GPUs as a benchmark. Any plain text corpus can be fed to this code.

TF with OpenBLAS on Saga

A corpus of 59 579 878 word tokens. Vocabulary size 10 000. 1 epoch.

Training time: 01:49:15

TF with OpenMKL on Saga

A corpus of 59 579 878 word tokens. Vocabulary size 10 000. 1 epoch.

Training time: 01:48:21