Search
Global Idle Computing Power Trains Model Rivaling R1 Performance
2025-05-13 /
Read about 0 minute
Author：小编
INTELLECT-2, the groundbreaking first distributed reinforcement learning (RL) training model in the world, has been officially unveiled. By leveraging global idle and decentralized computing resources, this model efficiently conducts RL training, drastically cutting down on training costs. Its performance stands on par with DeepSeek-R1, garnering substantial attention within the AI community. This pioneering achievement not only underscores the immense potential of RL training but also paves the way for more efficient AI model training in the future.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
