Collective Knowledge platform
1.3.1
Introduction
Introduction
Project overview
What is the Collective Knowledge project (CK)
Glossary
CK framework
CK portal
CK component
CK repository
Reproducible benchmarking pipeline and autotuning workflow
cBench
Live SOTA dashboards
Reproduced research papers
Portable CK solution
CK solutions for MLPerf
The meaning behind the ācā
Why CK?
Why open CK format?
Our goals and plans
Feedback and feature requests
Acknowledgments
Getting Started
cBench installation
Prerequisites
Linux
MacOS
Windows
Android (Linux host)
Docker
CK SDK installation
Prerequisites
Linux
MacOS
Windows
Android (Linux host)
Docker
MLPerf benchmark automation demo
User Guide
cBench commands
Setup and low-level access
Check the client version
Update client dependencies (CK components)
Setup the cKnowledge account
Test the login
Access the open cKnowledge API
Start the local server
CK components
Download components
Publish components
List versions of a given component
Open a cKnowledge web page with a given component
Delete components
cKnowledge dashboards
Initialize a cKnowledge graph
Push results
cKnowledge solutions
Download and initialize the existing solution
Run a solution
Benchmark a solution
Activate a virtual environment for the solution
List local solutions
Find local solutions
Delete local solutions
Create a new solution
CK commands
Practical use cases and tutorials
Share and reuse CK components similar to PyPI
Create customizable dashboards for live papers and collaborative experiments
Use cross-platform software detection plugins
Use cross-platform meta packages
Use portable workflows
Prepare portable cKnowledge solutions
How to contribute
Resources
CK Portal API
JSON Input
JSON Output
Available actions
Test login
List cKnowledge AI/ML/systems components
Download components
Get component info
Publish components
Get results
Publish results
cBench Python package
Submodules
cbench.client module
cbench.comm module
cbench.comm_min module
cbench.config module
cbench.graph module
cbench.main module
cbench.obj module
cbench.setup module
cbench.solution module
Module contents
CK Python package
Collective Knowledge platform
»
Index
Index
A
|
B
|
C
|
D
|
F
|
G
|
I
|
L
|
M
|
O
|
P
|
R
|
S
|
T
|
U
|
V
|
W
A
access() (in module cbench.comm)
activate() (in module cbench.solution)
B
benchmark() (in module cbench.solution)
C
cbench
module
cbench.client
module
cbench.comm
module
cbench.comm_min
module
cbench.config
module
cbench.graph
module
cbench.main
module
cbench.obj
module
cbench.setup
module
cbench.solution
module
D
delete() (in module cbench.obj)
deps_summary() (in module cbench.solution)
do_GET() (cbench.client.server_handler method)
do_POST() (cbench.client.server_handler method)
download() (in module cbench.obj)
download_file() (in module cbench.comm)
F
find() (in module cbench.solution)
G
get_platform_desc() (in module cbench.solution)
get_solution_dir() (in module cbench.solution)
get_work_dir() (in module cbench.config)
I
init() (in module cbench.graph)
(in module cbench.solution)
L
load() (in module cbench.config)
log_error() (cbench.client.server_handler method)
log_request() (cbench.client.server_handler method)
login() (in module cbench.setup)
ls() (in module cbench.solution)
M
module
cbench
cbench.client
cbench.comm
cbench.comm_min
cbench.config
cbench.graph
cbench.main
cbench.obj
cbench.setup
cbench.solution
O
open_page() (in module cbench.obj)
P
process_error() (in module cbench.main)
process_web_request() (in module cbench.client)
process_web_request_post_via_tunnel() (in module cbench.client)
publish() (in module cbench.obj)
publish_result() (in module cbench.solution)
push() (in module cbench.graph)
R
rm() (in module cbench.solution)
run() (in module cbench.solution)
S
send() (in module cbench.comm)
(in module cbench.comm_min)
server_handler (class in cbench.client)
setup() (in module cbench.setup)
start() (in module cbench.client)
T
ThreadedHTTPServer (class in cbench.client)
U
update() (in module cbench.config)
V
versions() (in module cbench.obj)
W
web_err() (in module cbench.client)
web_out() (in module cbench.client)