JUST ADD AI is involved in the development of generative AI solutions and custom language models. They have developed a secure, German-language GPT model and a specialized LLM for medical reports. Additionally, they develop custom language models that can be installed on-premise, indicating in-house training capabilities.
Companies training their own llms
Moreh is involved in the development and training of large language models (LLMs) as evidenced by their MoAI platform, which supports the training of multi-billion or multi-trillion parameter models. Additionally, they have trained the largest-ever Korean LLM with 221B parameters on their own infrastructure, indicating that they train these LLMs in-house rather than outsourcing the process.
Sola provides a solution for automating data entry and scraping using LLMs and computer vision. Additionally, the CTO has research experience in LLMs at the MIT-IBM Watson and Quest for Intelligence Labs, indicating in-house expertise in LLM development.
Jina AI has developed several tools and solutions related to AI and large language models (LLMs), including ThinkGPT and Jerboa, which are specifically mentioned as tools for augmenting and fine-tuning LLMs. Additionally, the company offers RunGPT, a framework for serving large multimodal models, and provides infrastructure for deploying and orchestrating search foundation models. These facts indicate that Jina AI is involved in the development and training of LLMs in-house.
NeuralMetrics is involved in the development and utilization of generative AI and Large Language Models (LLMs) for underwriting data solutions. They have launched a generative AI underwriting data workbench powered by generative AI and LLM support. Additionally, they have collaborated with Binghamton University to develop efficient quantum algorithms for training advanced language models, indicating that they are training these LLMs in-house rather than outsourcing the process.
OpenAI is a non-profit AI research company known for developing and training large language models (LLMs) such as GPT-3. The company conducts this training in-house, as evidenced by their pioneering research and development efforts in artificial general intelligence (AGI) and generative models.