LLM Models

WARNING: The checkpoints on this repo are not fully trained model. Evaluations of intermediary checkpoints and the final model will be added when conducted (see below).

"A parrot able to speak Japanese, ukiyoe, edo period" — Stable Diffusion XL

AquilaMoE: Efficient Training for MoE Models with Scale-Up and Scale-Out Strategies Language Foundation Model & Software Team Beijing Academy of Artificial Intelligence (BAAI) [Paper(released soon)] [Code] [github]

This is a 3B-parameter decoder-only language model with a focus on maximizing Japanese language modeling performance and Japanese downstream task performance. We conducted continued pretraining using Japanese data on the English language model, StableLM-3B-4E1T, to transfer the m…

State-of-the-art bilingual open-sourced Math reasoning LLMs. A solver , prover , verifier , augmentor .

Jamba2 Mini is an open source small language model built for enterprise reliability. With 12B active parameters (52B total), it delivers precise question answering without the computational overhead of reasoning models. The model's SSM-Transformer architecture provides a memory-e…

SEA-Guard is a collection of safety-focused Large Language Models (LLMs) designed specifically for the Southeast Asia (SEA) region. While the collection comprises four distinct models, we currently offer a single API endpoint that exclusively serves the Gemma-based model. You can…

SEA-LION-VL is an instruct-tuned vision-text model for the Southeast Asia (SEA) region.

An ablation of OctoCoder released for research purposes. Generally use OctoCoder, which performs better. Steps: 30

📖 Paper 🏠 Github 🤗 Spatial-SSRL-7B Model 🤗 Spatial-SSRL-3B Model 🤗 Spatial-SSRL-Qwen3VL-4B Model 🤗 Spatial-SSRL-81k Dataset 📰 Daily Paper

- Arxiv - Github - Model Collection - Data

Medical-Qwen3-Swallow-32B is a medical-domain language model based on tokyotech-llm/Qwen3-Swallow-32B-RL-v0.2. It is designed to support research and development toward safe and trustworthy AI for Japanese clinical settings.