view article Article Introducing harness adapters: adapt your favorite model to your harness to improve performance and cost +8 ybelkada, burtenshaw, SaylorTwift, lysandre, aminediroHF, AdithyaSK, evalstate, sergiopaniego, joelniklaus, lewtun • 1 day ago • 38
view article Article Welcome Inkling 2 Mini, Cutting-edge Performance for your Local Setup +2 merve, pcuenq, ariG23498, guarin • 1 day ago • 44
thinkingmachines/Inkling-2-Mini Image-Text-to-Text • 26B • Updated about 18 hours ago • 7.94k • • 1.57k
view post Post 223 Run Inkling 2 Mini, the strongest 26.4B model that runs on 16GB RAM.Thinking Machines’ Apache 2.0 MoE model uses a looped transformer arch & scores 48.2% on DeepSWE v1.1, the best for size.Run on Unsloth Desktop or llama.cpp.GGUF: unsloth/Inkling-2-Mini-GGUFGuide: https://unsloth.ai/docs/models/inkling-2-mini See translation 🚀 1 1 ❤️ 1 1 🔥 1 1 + Reply
Energy-Based Reasoners: Your Recurrent Reasoner Should Be an Energy-Based Model Paper • 2611.08977 • Published 16 days ago • 11