HF's Merve Noyan shares beginner guides for zero-shot classifiers, says many LLM tasks never needed them
ceciletamura · x · 2026-09-21
Hugging Face's Merve Noyan shared a set of beginner-friendly guides for text classifiers, covering both standard fine-tuning (DeBERTa, ModernBERT) and zero-shot classification, plus a multimodal image-text example, with links to relevant models on the Hub.
In her accompanying commentary she quipped that people comparing small classifiers against GPT-5.6 have clearly never fine-tuned a BERTForXYZ for a living — zero-shot classifiers are fascinating but were never scaled the way decoder-only models were, and many problems now solved with LLMs could have been solved with them. 'It was a skill issue.'
More from Models
- GLM 5.3 Flash, DeepSeek V4.1 Flash and Qwen 3.8 all beat 'frontier' models from just 10 months ago — burny_tech · 2026-09-21
- Reddit meme: Grok 'devolves' while local MiniMax 3 wins users over — Mystvearn_ · 2026-09-21
- Two years after 'intelligence too cheap to meter', $10/$50 models are the new norm — teortaxesTex · 2026-09-21
- Gary Marcus: "General" LLMs Fall Apart the Moment They Leave Their Training Distribution — GaryMarcus · 2026-09-21
- AI hallucinated a passenger's middle name, almost costing him his flight — menhguin · 2026-09-21
- Distilled Qwen3.8-35B-A3B GGUF quantized model trends on Hugging Face — empero-ai · 2026-09-21