Platform
The Platform comprises the hard- and software underlying the actual Implementation. The metadata information about the Platform should include the underlying hardware, the operating system, and the used software libraries. software
covers every package or library that is used for the Implementation. If a retrieval toolkit is used, it should be reported explicitly.
Checklist
platform
→hardware
→cpu
→model
Description: Name of the CPU model.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→hardware
→cpu
→architecture
Description: Identifier of the CPU architecture.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→hardware
→cpu
→operation mode
Description: Operation mode of the CPU.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→hardware
→cpu
→number of cores
Description: Number of CPU cores.
Type: Scalar
Encoding: A decimal integer number;!!int
.platform
→hardware
→gpu
→architecture
Description: Name of the GPU architecture.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→hardware
→gpu
→number of cores
Description: Number of GPU cores.
Type: Scalar
Encoding: A decimal integer number;!!int
.platform
→hardware
→gpu
→memory
Description: Amount of available memory of the GPU; string with numbers followed by GB.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→hardware
→ram
Description: Amount of available RAM; string with numbers followed by GB.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→operating system
→kernel
Description: The kernel version of the operating system.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→operating system
→distribution
Description: The name of the operating system's distribution.
Type: Scalar
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→software
→libraries
Description: Names and versions of the software libraries and packages underlying the experiment's Implementation with the following syntax<libray-name>==<version>
. If possible, libraries and packages of different programming languages should be in separate nodes (see example below).
Type: Sequence of scalars;!!seq
.
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.platform
→software
→retrieval toolkit
Description: Names and versions of the retrieval toolkits underlying the experiment's Implementation with the following syntax<toolkit-name>==<version>
.
Type: Sequence of scalars;!!seq
.
Encoding: UTF-8 encoded string of characters (RFC3629);!!str
.
Naming convention:indri
,terrier
,anserini
,pyserini
,pyterrier
,solr
,elasticsearch
Example
platform:
hardware:
cpu:
model: Intel(R) Xeon(R) Gold 6144 CPU @ 3.50GHz
architecture: x86_64
operation mode: 64-bit
number of cores: 16
gpu:
model: NVIDIA RTX A6000
memory: 48 GB
number of cores: 10752
ram: 32 GB
operating system:
kernel: GNU/Linux 4.15.0-166-generic
distribution: Ubuntu 18.04.5 LTS
software:
libraries:
python:
- blas==1.0
- libgfortran==3.0.1
- libxml2==2.9.8
- lightgbm==2.2.1
- ncurses==6.1
- numpy==1.15.4
- numpy-base==1.15.4
- scikit-learn==0.20.1
- scipy==1.1.0
- setuptools==40.6.2
java:
- lucene==7.6
retrieval toolkit:
- anserini==0.3.0