Qwen3.8-27B coding-agent-adapted model trends on Hugging Face
harness-hub · hf · 2026-11-19
harness-hub's Qwen3.8-27B-opencode-adapted is trending on Hugging Face as an image-text-to-text model.
It was fine-tuned with LoRA and (async) GRPO, adapted for the opencode coding agent workflow, with tags spanning agents, harbor, and openenv.
More from coding & agent
- Agno's agent compaction: summary as index, originals stay searchable — pritisinghhhh · 2026-11-19
- "Copy your AGENTS.md if AI coding left you zero problems to solve" — burny_tech · 2026-11-19
- Dev open-sources Traffic Engine: DOM playbook cache makes repeat browser agent sessions cost $0 tokens — ystudio3d · 2026-11-19
- Hugging Face team preps docs to build, package and distribute Helion kernels on the Hub — ariG23498 · 2026-11-19
- Developer Adds a Dynamic Island to His App in 30 Minutes With Just 3 Prompts, No Swift Needed — PrajwalTomar_ · 2026-11-19
- Ditching agent frameworks made my agents noticeably better: eval and side-effect checks are the real work — Rbovzz · 2026-11-19