IBM, the technology giant, has announced the addition of new generative AI models and capabilities to its Watsonx data science platform. As part of the update, IBM is introducing the Granite series models, which are large language models (LLMs) similar to OpenAI’s GPT-4 and ChatGPT. These models are designed to summarize, analyze, and generate text.
Key Takeaway
IBM is bolstering its Watsonx data science platform by introducing new generative AI models and capabilities, including the Granite series models. These models utilize high-quality curated data to provide specialized domain-specific models that support various NLP tasks. With enhancements in customizability, synthetic data generation, and data discovery, IBM aims to support clients in scaling AI adoption in a secure and trustworthy manner.
The Power of IBM’s Granite Series Models
The Granite series models developed by IBM have been trained using curated and high-quality enterprise data, instead of publicly scraped data. These models offer specialized subsets within different domains, such as finance. This enables AI builders to use smaller, domain-specific models that can perform as well as larger general models. The Granite series models support a wide range of natural language processing (NLP) tasks, including summarization, content generation, and insight extraction.
Tailoring AI Models with Tuning Studio
IBM is also introducing Tuning Studio in its Watsonx.ai component, which allows users to customize generative AI models according to their specific data needs. By using Tuning Studio, customers can fine-tune models with as few as 100 to 1,000 examples. After specifying the task and providing labeled examples, the model can be deployed via an API from the IBM Cloud.
Synthetic Data Generator for Tabular Data
Furthermore, IBM is set to introduce a synthetic data generator for tabular data in Watsonx.ai. This tool enables companies to generate synthetic data from custom data schemas and internal datasets. The aim is to extract insights for AI model training and fine-tuning, while reducing risk. However, it’s important to note that the exact meaning of “reduced risk” has not been explained by IBM.
Enhancements in Watsonx.data and Watsonx.governance
IBM is also making enhancements to Watsonx.data, its data store that enables users to access data while applying query engines, governance, automation, and integrations with existing databases and tools. In the future, customers will have access to a self-service, chatbot-like tool that simplifies data discovery, augmentation, visualization, and refinement for AI applications.
Additionally, IBM plans to introduce vector database capabilities to support retrieval-augmented generation (RAG) in Watsonx.data. RAG is an AI framework that improves the quality of LLM-generated responses by grounding the model on external knowledge sources.
IBM is also launching Watsonx.governance, a toolkit that provides mechanisms to protect customer privacy, detect model bias and drift, and help organizations meet ethics standards.
IBM’s Focus on Supporting Clients and Scaling AI
IBM understands the competitive AI landscape and the need to demonstrate its capabilities. The company’s CEO, Arvind Krishna, has stated that AI plays a crucial role in IBM’s future growth. IBM’s hybrid cloud and AI technologies, including Watsonx, are witnessing healthy adoption rates, with over 150 corporate customers already utilizing the platform. Partnerships with major companies, such as Samsung and Citi, underscore the growing interest in IBM’s AI offerings.
