Qwen 2.5

New 1.5B router model achieves 93% accuracy without costly retraining ChatGPT-Image-Jul-2-2025-09_19_17-AM.png?w=578

admin July 7, 2025

Katanemo Labs’ new LLM routing framework aligns with human preferences and adapts to new models without retraining.Read...

Qwen swings for a double with 2.5-Omni-3B model that runs on consumer PCs, laptops cfr0z3n_vector_art_graphic_novel_style_image_of_a_humanoid_robo_bbc90f34-b66b-4b30-8405-c858eb5122a9.png?w=578

admin April 30, 2025

The Qwen2.5-Omni-3B model is licensed for non-commercial use only under Alibaba Cloud’s Qwen Research License Agreement.Read More

Former DeepSeeker and collaborators release new method for training reliable AI agents: RAGEN nuneybits_Vector_art_of_college_classroom_with_robots_in_it_a5945ebd-0047-40a5-aeb7-79daba2024a0.webp?w=578

admin April 23, 2025

RAGEN stands out not just as a technical contribution but as a conceptual step toward more autonomous,...

Alibaba’s new open source model QwQ-32B matches DeepSeek-R1 with way smaller compute requirements cfr0z3n_minimalist_corporate_memphis_flat_illustration_a_robo_0132573d-abf5-4d21-aedd-5cc6b6f17164_3.png?w=578

admin March 6, 2025

While DeepSeek-R1 operates with 671 billion parameters, QwQ-32B achieves comparable performance with a much smaller footprint.Read More

New open-source math model Light-R1-32B surpasses equivalent DeepSeek performance with only $1000 in training costs cfr0z3n_humanoid_robot_wearing_crown_teaches_advanced_math_wi_99a27404-b5e3-4e99-8f76-dd1865a3282f_1.png?w=578

admin March 6, 2025

Companies can freely deploy Light-R1-32B in commercial products, maintaining full control over their innovations.Read More

Researchers find you don’t need a ton of data to train LLMs for reasoning tasks an-illustration-of-a-3d-sphere-with-connected-data-lNRjO3frQBCyMol5ef96OQ-s5VN3KT1TDOXoBGddLHd6Q.jpeg?w=578

admin February 14, 2025

With a few hundred well-curated examples, an LLM can be trained for complex reasoning tasks that previously...

You may have missed