lightblue/ao-karasu-72B

The lightblue/ao-karasu-72B is a 72.3 billion parameter causal language model developed by lightblue, featuring a 32768-token context length. This model is specifically trained on a diverse Japanese dataset, including Wikipedia-based QA, technical blogs, Japanese QA site answers, LLM-generated prompts, and news articles. It is optimized for Japanese language understanding and generation, making it suitable for applications requiring high-quality Japanese text processing.

Cold
Public
72.3B
FP8
32768
Hugging Face