Hugging Face's transformers serve can turn 250 LLMs & VLMs into logit-based decision models
LysandreJik · x · 2026-10-12
Lysandre Dubreuil (transformers maintainer) announced a fun addition to transformers serve: a new /v1/systemone endpoint scores decisions using the model's logits — no text generation needed — turning any of 250 LLM & VLM models into a JEV-like decision model, demoed with a microducks race.
More from coding & agent
- LightOn ships 2B code-search agent ColGrep with model, GGUF and paper — IgorCarron · 2026-10-12
- LightOn's 2B search agent matches 27B open-source models on adapted SWE-bench evals — IgorCarron · 2026-10-12
- ColGrep-Agent cuts per-question tool calls from 7.1 to 4.0 by searching locally — IgorCarron · 2026-10-12
- ColGrep-Agent cuts Claude Code API tokens 34.3% and cost from 7.3¢ to 4.4¢ per question — IgorCarron · 2026-10-12
- Benchmarking Claude Code search tools: grep vs ColGrep on 52 code-search tasks — IgorCarron · 2026-10-12
- ColGrep-Agent: a local search loop in three commands, 5.5s per query on H100 — IgorCarron · 2026-10-12