Hermes LoRA Adapter Released for LFM2.5-24B, Cutting Tool Calls 20%

harness-hub has released a Hermes agent LoRA adapter for LiquidAI's LFM2.5-24B-A2B that cuts tool calls by 20% with no accuracy loss on held-out tasks, along with a 37MB GGUF version loadable directly in llama.cpp.

2026-11-02 ~ 2026-11-02 · 2 related posts

• RL-trained LFM2.5-24B adapter matches accuracy with 20% fewer tool calls — yb2698 · 2026-11-02
• LFM2.5-24B Hermes adapter ships as GGUF: local agents with 20% fewer tool calls — yb2698 · 2026-11-02