Solar Open 2 grows to 250B parameters while cutting total training tokens
keunwoochoi · x · 2026-07-26
- Solar Open 2 is described as a 250B sovereign LLM from South Korea, and the author says it is 2.5× larger than Solar Open 1 while using fewer total training tokens.
- The post highlights a four-stage post-training recipe: supervised fine-tuning, multi-domain RL, specialist training, and multi-teacher on-policy distillation.
- The paper also claims selective weight transfer from Solar Open 1, a 10T-token general pre-training stage, a 1T intensive stage, and Korean officework data curation for domain-specific performance.
Related event: Upstage Releases Solar Open 2: A 250B Sovereign LLM(6 posts)→
More from Models
- GPT-6 Astra beats Factorio with enemies in 44 in-game hours at ~$4,500 API cost — liminal_bardo · 2026-09-11
- awesome-llm-leaderboards: an open-source directory of LLM leaderboards, pricing tables, comparison tools — Last_Establishment_1 · 2026-09-11
- Anthropic claims it works to keep eval environments unidentifiable to models — MaxKannen · 2026-09-11
- Nex N2.5 Pro released on Hugging Face with 407GB of weights — jinnyjuice · 2026-09-11
- RoMa v2 image matching model unveiled in the usual black poster — ducha_aiki · 2026-09-11
- OpenAI rated Astra 'Critical' for cyber capabilities — and admits it's harder to monitor — theguywhobuilds · 2026-09-11