Introducing Hy4 Preview
Introducing Hy4 Preview New open weight text input (no vision) LLM from Chinese company Tencent today: 770B total parameters, 49B active parameters, 1M token context window, 1.56TB on Hugging Face. This is a big size increase from their previous Hy3 in July, which was 295B, 21B
The unveiling of Hy4 Preview by Tencent marks a significant milestone in the development of large language models (LLMs). With 770B total parameters and 49B active parameters, this new model represents a substantial leap forward from its predecessor, Hy3, which had 295B total parameters and 21B active parameters. This increase in size and capacity enables Hy4 to handle more complex and nuanced tasks, potentially leading to improved performance in various natural language processing applications.
The expanded context window of 1M tokens is also noteworthy, as it allows the model to consider longer sequences of text and relationships between them. This feature can be particularly beneficial for tasks such as text summarization, question answering, and language translation. Furthermore, the release of Hy4 Preview on Hugging Face, a popular platform for sharing and deploying AI models, makes it easily accessible to developers and researchers, facilitating further innovation and experimentation.
As the AI landscape continues to evolve, it's essential to monitor the performance and applications of Hy4 Preview. Key areas to watch include its impact on downstream tasks, such as chatbots, content generation, and language understanding. Additionally, comparing Hy4's capabilities with those of other state-of-the-art LLMs, like those from Google, Meta, or Microsoft, will provide valuable insights into Tencent's advancements and the overall progress of the field.
Originally reported by simonwillison.net. WebNews adds analysis for ai & agent economy readers.