With Insilico, we got our LFM 2.6B model customized on retrosynthesis and it performs better than Astra! ๐๐๐
This is huge albeit expected improvement from @OpenAI for GPT-6 Astra model. Great work, best LLM out there!
What is eye-catching - it outperformed two strong specialists models!
Notable that our 2.6B specialist is still the best on single-step retrosynthesis ๐#insilicoSOTAFM#
๋ ๋ณด๊ธฐ
