ai-and/termgrade-gemma4-31b-lora - Каталог нейросетей
Генерация текста

ai-and/termgrade-gemma4-31b-lora

Добавлено:
ai-and/termgrade-gemma4-31b-lora

Part of TermGrade: graded environments and trajectories for terminal agents. Read the blog post. 46.1 on Terminal-Bench 2.1 · +3.1 over base · 151 training environments A LoRA adapter for google/gemma-4-31B-it, RL post-trained with GRPO++ on terminal tasks selected by their measured base pass rate. Pre-merged weights: ai-and/termgrade-gemma4-31b-merged · training data: ai-and/termgrade-bandcentral-gemma4-31b-bash 46.1 is the mean of seven evaluation runs on this checkpoint (four on the adapter, three on the merged weights), at three attempts per task each, so 21 attempts per task. We trained the identical recipe five times end to end and evaluated every result: t = 3.8 on four degrees of freedom, p ≈ 0.025. +2.1 is the mean of the five runs; +3.1 is this particular checkpoint, which sits at the top of the range. There is no fixed random seed, so the five are independent replicates of one configuration. > † 43.0 is Artificial Analysis’ published figure. > Our own reproduction of the base came out at 42.7; we quote against 43.0 because using ours > would inflate every gain by 0.3. Their number is one three-repeat evaluation. One benchmark, five runs. We report the five-run average…

Модальности:
Генерация текста


Задача: Генерация текста
Автор: ai-and
Теги: peft, lora, terminal-bench, agentic, rl, grpo, dapo, verl
Лайков: 3  |  Загрузок: 79

Открыть на HuggingFace →

Описание основано на материалах HuggingFace. Перевод выполнен автоматически.