Legal Notice
Last update : 02-22-2022
Publisher
The user is informed that the website https://www.datakalab.com/ (the "Website") is published by the company:
DATAKALAB SAS7 passage du Puits Bertin, 92110 ClichySIRET N° 81883027500017With a share capital of €1066.00, registered with the Nanterre RCS under no. 818 830 275.
Director of publication: Mr Lucas Fischer.
Contact
The User has the possibility to contact Datakalab :
● By simple letter to the above address
● By e-mail to moc.balakatad%40olleh
Website hosting
JETIMPEX, INC. - affiliated project «Weblium»525 NE 14 Avenue,Fort Lauderdale,FL 33301, United States.
Conditions of access to the Website
Access to the Website is free of charge.Differences may exist between the computer version of the website and its mobile version.Access to the Website as well as the use of its content is made within the framework of the legal notices described below. The fact of accessing and browsing the Website constitutes your unreserved acceptance of these conditions.Datakalab informs you that the present mentions may be modified at any time. These modifications are published by being put online and are deemed to be accepted without reservation when you access the Website after they have been put online. We recommend that you consult this page regularly.
Warranty and liability
Datakalab makes every effort to ensure the best possible availability of the Website and the accuracy and updating of its content, provided for information purposes. However, Datakalab makes no commitment of availability and does not guarantee that the Website is free of errors or omissions. Datakalab is free to modify, at any time and without notice, the content of the Website.The user undertakes to use the Website and all the information and documents to which he or she may have access via the Website only for purposes that comply with public order, morality and the rights of third parties.Datakalab declines all responsibility in the event of interruption or inaccessibility of the Website, the occurrence of bugs, any damage resulting from fraudulent acts of third parties from the Website.Datakalab cannot be held liable for any direct or indirect damage, whatever the cause, origin, nature and consequence, resulting from the consultation or use of the Website.
Intellectual Property
The Website as a whole and each of the elements that make it up (such as texts, articles, photographs, illustrations, diagrams, logos, trademarks, architecture, software) are protected by intellectual property rights and are the exclusive property of Datakalab or the property of a third party that has licensed this content to Datakalab. Visiting the Website does not constitute a transfer of rights of any kind whatsoever. The use of all or part of the Website and its elements, in particular by downloading, reproduction, transmission, representation or distribution for purposes other than for your personal and private use for non-commercial purposes is prohibited.
Links to third party sites
www.datakalab.com points to third party sites. These target sites are not under the control of Datakalab, which is therefore not responsible for the content of these sites, the links they contain or the changes or updates made to them.
Personal data
Datakalab may collect and process personal data in the context of the use of the Website.To learn more about how Datakalab processes and protects your personal data, you can consult our Privacy Policy.
Applicable law
The Website and these terms of use are governed by French law. In the event of a dispute relating to the Website and in the event that an amicable agreement cannot be reached, the competent court will be determined according to the applicable rules of procedure.
How do you get reliable AI on the edge when it was made for the cloud? With millions or billions of parameters to process, today’s deep neural nets are too big for CPUs and MCUs to process in real time .
Until now...
Say hello to embedded AI that is fast, frugal and accurate.
Context Adaptation, Quantization, Pruning and Batch Norm Folding formodels that are up to 330x smaller and 20x times faster.