Kimi k1.5 is an advanced multi-modal model trained with reinforcement learning (RL) that achieves state-of-the-art reasoning performance across multiple benchmarks and modalities. It outperforms existing models such as GPT-4o and Claude Sonnet 3.5 in short-CoT tasks by a large margin and matches OpenAI's o1 in long-CoT performance.
Kimi k1.5 is an o1-level multi-modal model trained with reinforcement learning (RL) that achieves state-of-the-art reasoning performance across multiple benchmarks and modalities. It outperforms existing models such as GPT-4o and Claude Sonnet 3.5 in short-CoT tasks by a large margin and matches OpenAI's o1 in long-CoT performance. Key ingredients of Kimi k1.5 include long context scaling, improved policy optimization, a simplistic RL framework, and multi-modal training.
The Emotional First Aid Raw Dataset is a collection of raw, unannotated psychological counseling Q&A data, designed to support research in AI applications for mental health. It contains over 172,000 topics with 2,381,273 messages, totaling 44,514,786 characters, providing a rich source of data for natural language processing and AI development.
We're on a journey to advance and democratize artificial intelligence through open source and open science.
FriendnPal is revolutionizing mental health support in Africa as the first AI-powered app with multilingual capabilities, reflecting our commitment to inclusivity and accessibility, ensuring that support is available to everyone, regardless of language barriers.. At FriendnPal, our dedicated team is driven by a shared passion for mental health advocacy and innovation. With diverse backgrounds and expertise, we're united in our mission to make a positive impact on the lives of individuals across Africa. Together, we're committed to harnessing the power of technology and compassion to create a platform that empowers individuals to prioritize their mental well-being.