340M open model matches fine-tuned Jev on two clinical tasks, zero-shot on CPU
MaziyarPanahi · x · 2026-09-29
Developer Maziyar Panahi reports that fastinoAI's open GLiNER2.5-Decide (340M params) ran zero-shot on a CPU — with zero fine-tuning — and scored a perfect 12/12 on two clinical NLP tasks (assertion status and allergy documentation), matching the fine-tuned model Jev.
The full benchmark stands at 36/48 versus Jev's 48/48, but the zero-shot result is striking. He's inviting others to fine-tune it on clinical notes to close the remaining gap.
Related event: 340M open-source model aces 12 clinical tasks zero-shot(2 posts)→
More from Models
- Opus 5.5 dramatically cuts em-dash usage, blurring AI writing detection — jonathan_wilke · 2026-09-29
- PrivacyBench v2 launches: micro1's flow-transform 1.0 leads at 95.84%, 9.64 points ahead — Exp_Mark · 2026-09-29
- Grok 4.7 xHigh Tops Artificial Analysis Cyber Index for Enterprise Cyber Defense — XFreeze · 2026-09-29
- Dev Re-Verified Benchmarks Repeatedly: New Model Complements Opus 5.5 — giansegato · 2026-09-29
- Anthropic exec on Sonnet 5.5's design skills, teases Haiku 5.5 within weeks — mikeyk · 2026-09-29
- Leak claims OpenAI GPT-6 sol brutally outclassed by Claude Sonnet 5.5 — ns123abc · 2026-09-29