Known large language models (LLMs)

As of today, October 26, 2023, here’s a list of some of the currently known large language models (LLMs):

Open-Source:

  • BERT (Bidirectional Encoder Representations from Transformers): Developed by Google AI, BERT is a foundational LLM trained on a massive dataset of text and code.
  • GPT-2 and GPT-3 (Generative Pre-trained Transformer): Developed by OpenAI, GPT-2 and GPT-3 are known for their ability to generate creative text formats and respond to diverse prompts.
  • Jurassic-1 Jumbo: Developed by AI21 Labs, Jurassic-1 Jumbo is a factual language model trained on a massive dataset of text and code, known for its factual accuracy and information retrieval capabilities.
  • WuDao 2.0: Developed by the Beijing Academy of Artificial Intelligence, WuDao 2.0 is a Chinese LLM trained on a massive dataset of Chinese text and code.
  • Megatron-Turing NLG: Developed by NVIDIA and Microsoft, Megatron-Turing NLG is a factual language model known for its ability to generate different creative text formats and answer your questions in an informative way.

Commercial:

  • LaMDA (Language Model for Dialogue Applications): Developed by Google AI, LaMDA is a conversational LLM used in several Google products like Search and Assistant.
  • Megatron-Turing NLG (also mentioned in Open-Source): While released as open-source, NVIDIA also offers a commercially licensed version with additional features and support.
  • WuDao 2.0 Pro: Developed by the Beijing Academy of Artificial Intelligence, WuDao 2.0 Pro is a commercially licensed version of WuDao 2.0 with extended capabilities.
  • Bard (me!): Developed by Google AI, Bard is a factual language model currently under development, designed to be informative and comprehensive.

Please note:

  • This list is not exhaustive, and new LLMs are constantly being developed.
  • The specific capabilities and limitations of each LLM vary depending on its training data, architecture, and intended use.
  • Some LLMs may be more readily available for research or public use than others.

We will keep you posted. I hope this information is helpful!


Discover more from QUE.com

Subscribe to get the latest posts sent to your email.

Founder & CEO, EM @QUE.COM

Founder, QUE.COM Artificial Intelligence and Machine Learning. Founder, Yehey.com a Shout for Joy! MAJ.COM Management of Assets and Joint Ventures. More at KING.NET Ideas to Life | Network of Innovation

kingdotnet has 2745 posts and counting.See all posts by kingdotnet

Leave a Reply

Connect with

Discover more from QUE.com

Subscribe now to keep reading and get access to the full archive.

Continue reading