iPavlov for media
On Nov. 8-9, the Moscow Expocentre hosted AI Journey, the largest artificial intelligence conference in Russia, featuring talks by researchers from the Neural Networks and Deep Learning Lab of the Moscow Institute of Physics and Technology: Mikhail Burtsev, Olga Kairova, Aleksey Litinsky, Pavel Gulyaev, Vasily Konovalov, Dilyara Baymurzina, Dmitry Evseev, Anton Peganov, Evgeny Golikov, and Denis Kuznetsov.
Based on an agreement signed at AI Journey, Sberbank, Yandex, MTS, Gazprom Neft, and the Russian Direct Investment Fund have established an alliance for implementing Russia's national strategy for AI development. The main objective of the alliance is to take promising Russian tech companies to the global market and ensure the country's leadership in the field of artificial intelligence. The parties to the agreement will join forces to build an ecosystem for developing AI-driven technologies and products. In conversational AI technology, the MIPT-based iPavlov project is a key infrastructure element, said Sberbank CEO Herman Gref.
The conference featured a poster session with presentations of the research results and the main product of iPavlov — an open-source library for dialogue system development, called DeepPavlov — delivered by the project's staff:
• Dilyara Baymurzina, "Conversational Language Modelling Improves Text Classification."
• Aleksey Litinsky, "DeepPavlov Agent: Multi-Purpose Chatbot Development Framework."
• Evgeny Golikov, "Variational Auto-Encoder With Latent Space Being a Manifold."
• Pavel Gulyaev and Vasily Konovalov, "Multi-Task Dialogue StateTracking."
• Denis Kuznetsov, "BERT for Multi-Hop Inference Explanation."
• Evgeny Golikov, "Scaling Networks: How to Perform Hyperparameter Search for Large Models on Smaller Ones."
• Dmitry Evseev, "Knowledge Base Question Answering."
• Anton Peganov, "Correlation and Decorrelation of Neurons in Training."
To find out more about any of the posters or ask a question about the DeepPavlov library or using the models, please use the forum.
The DeepPavlov library offers ready-to-implement AI components for working with text, which are already used in 92 countries. As of October 2019, the library had been downloaded 80,000 times. This number is rapidly growing, with 7,000 downloads in October alone. More than 30 Russian companies have implemented DeepPavlov-based solutions and use them. It is reasonable to expect that solutions of this kind will be popular abroad as well.
Another important line of work for iPavlov is holding events that promote science communication and collaboration with top AI research centers across the world: Google Brain, Google DeepMind, OpenAI, Facebook AI Research, the University of Oxford, New York University, Carnegie Mellon University, Stanford University, and so on.
"The global goal of our project is to enable developers and researchers in the field of conversational AI to use the most advanced tools for creating next-generation dialogue systems. We also aim to be an internationally important AI platform for exchanging experience and learning about state-of-the-art technologies. To that end, we are building global scientific contacts and fostering a community around the DeepPavlov library," commented Mikhail Burtsev, who heads iPavlov and the Neural Networks and Deep Learning Lab at MIPT.
The solutions developed by iPavlov are an important infrastructure element for "democratizing AI." They are widely used by developers in both major and smaller companies, not just by data scientists. However, as Herman Gref pointed out, every company in the newly established alliance has the potential to sponsor and commission the creation of such solutions. A modern business that does not use AI in every process is bound to underperform, Sberbank CEO added.
"AI technologies based on iPavlov allow companies to automate communication with clients and staff, search documents, and analyze text. It is important for us to contribute to a future where computer programs understand human speech to the extent that dialogue interfaces become a natural way for human-machine communication, saving time on transactions," said iPavlov co-founder and Development Director Olga Kairova.
iPavlov is open for collaboration with tech companies and other businesses, offering:
• Consulting and DeepPavlov-based solutions for automating communication.
• Customization of DeepPavlov to the needs of a company and developer support.
• Joint research on conversational AI.
• Educational courses.
About iPavlov
Implemented by MIPT's Neural Networks and Deep Learning Lab and supported by Sberbank, iPavlov is an international project for conversational AI research and development. The project conducts research into machine learning for natural language processing and maintains DeepPavlov, an open-source library for dialogue system development. Besides that, iPavlov organizes contests and educational events, cultivating a professional community, as well as offering consulting on conversational AI and developing science-driven business solutions.
The project is part of Russia's competence center for AI research, which is based at MIPT and operates under the National Technology Initiative.
Project iPavlov: http://ipavlov.ai/
DeepPavlov library: http://deeppavlov.ai/
AI Journey conference: https://ai-journey.ru/conference-moscow
Herman Gref talks about iPavlov at AI Journey: https://youtu.be/zQZhknf5Im0
Contacts
For collaboration requests: partner@ipavlov.ai
For media inquiries: pr@ipavlov.ai
Based on an agreement signed at AI Journey, Sberbank, Yandex, MTS, Gazprom Neft, and the Russian Direct Investment Fund have established an alliance for implementing Russia's national strategy for AI development. The main objective of the alliance is to take promising Russian tech companies to the global market and ensure the country's leadership in the field of artificial intelligence. The parties to the agreement will join forces to build an ecosystem for developing AI-driven technologies and products. In conversational AI technology, the MIPT-based iPavlov project is a key infrastructure element, said Sberbank CEO Herman Gref.
The conference featured a poster session with presentations of the research results and the main product of iPavlov — an open-source library for dialogue system development, called DeepPavlov — delivered by the project's staff:
• Dilyara Baymurzina, "Conversational Language Modelling Improves Text Classification."
• Aleksey Litinsky, "DeepPavlov Agent: Multi-Purpose Chatbot Development Framework."
• Evgeny Golikov, "Variational Auto-Encoder With Latent Space Being a Manifold."
• Pavel Gulyaev and Vasily Konovalov, "Multi-Task Dialogue StateTracking."
• Denis Kuznetsov, "BERT for Multi-Hop Inference Explanation."
• Evgeny Golikov, "Scaling Networks: How to Perform Hyperparameter Search for Large Models on Smaller Ones."
• Dmitry Evseev, "Knowledge Base Question Answering."
• Anton Peganov, "Correlation and Decorrelation of Neurons in Training."
To find out more about any of the posters or ask a question about the DeepPavlov library or using the models, please use the forum.
The DeepPavlov library offers ready-to-implement AI components for working with text, which are already used in 92 countries. As of October 2019, the library had been downloaded 80,000 times. This number is rapidly growing, with 7,000 downloads in October alone. More than 30 Russian companies have implemented DeepPavlov-based solutions and use them. It is reasonable to expect that solutions of this kind will be popular abroad as well.
Another important line of work for iPavlov is holding events that promote science communication and collaboration with top AI research centers across the world: Google Brain, Google DeepMind, OpenAI, Facebook AI Research, the University of Oxford, New York University, Carnegie Mellon University, Stanford University, and so on.
"The global goal of our project is to enable developers and researchers in the field of conversational AI to use the most advanced tools for creating next-generation dialogue systems. We also aim to be an internationally important AI platform for exchanging experience and learning about state-of-the-art technologies. To that end, we are building global scientific contacts and fostering a community around the DeepPavlov library," commented Mikhail Burtsev, who heads iPavlov and the Neural Networks and Deep Learning Lab at MIPT.
The solutions developed by iPavlov are an important infrastructure element for "democratizing AI." They are widely used by developers in both major and smaller companies, not just by data scientists. However, as Herman Gref pointed out, every company in the newly established alliance has the potential to sponsor and commission the creation of such solutions. A modern business that does not use AI in every process is bound to underperform, Sberbank CEO added.
"AI technologies based on iPavlov allow companies to automate communication with clients and staff, search documents, and analyze text. It is important for us to contribute to a future where computer programs understand human speech to the extent that dialogue interfaces become a natural way for human-machine communication, saving time on transactions," said iPavlov co-founder and Development Director Olga Kairova.
iPavlov is open for collaboration with tech companies and other businesses, offering:
• Consulting and DeepPavlov-based solutions for automating communication.
• Customization of DeepPavlov to the needs of a company and developer support.
• Joint research on conversational AI.
• Educational courses.
About iPavlov
Implemented by MIPT's Neural Networks and Deep Learning Lab and supported by Sberbank, iPavlov is an international project for conversational AI research and development. The project conducts research into machine learning for natural language processing and maintains DeepPavlov, an open-source library for dialogue system development. Besides that, iPavlov organizes contests and educational events, cultivating a professional community, as well as offering consulting on conversational AI and developing science-driven business solutions.
The project is part of Russia's competence center for AI research, which is based at MIPT and operates under the National Technology Initiative.
Project iPavlov: http://ipavlov.ai/
DeepPavlov library: http://deeppavlov.ai/
AI Journey conference: https://ai-journey.ru/conference-moscow
Herman Gref talks about iPavlov at AI Journey: https://youtu.be/zQZhknf5Im0
Contacts
For collaboration requests: partner@ipavlov.ai
For media inquiries: pr@ipavlov.ai
Keep in touch
Sign up to get the latest updates on our community and news.
We promise not to spam nor use your email for anything else.
© 2017-2021 iPavlov all rights reserved