A full fine-tune of LiquidAI/LFM2.5-2.6B for agentic data-analysis tasks, trained with asynchronous GRPO (TRL Async GRPO) using OpenCode, Claude Code, Codex, Mini-SWE-Agent. This release is step 1,000 on main, scoring 54.2% pass@1 across four evaluation harnesses. Article · Collection · Evaluation tasks · Training dashboard Training curves use a trailing 50-step mean (at least 10 observations). Evaluation markers show measured checkpoints; hollow markers and dotted segments indicate incomplete coverage. Faint background lines show the full recorded trajectory; colored lines reveal the measured checkpoints. The first frame shows the completed chart before replaying, so previews also contain the full curves. Tool-call savings compare tasks solved by both the base model and checkpoint, with a different matched cohort at each checkpoint. The animation stops at this revision’s step 1,000. TRL Async GRPO with correctness plus a tool-call efficiency bonus, initialized from the base model. Each group uses one task and one harness; harnesses rotate across groups. Rollouts run through OpenEnv × Harbor in E2B sandboxes. The reward is correctness × (1 + 0.1 × 15 / (15 + toolcalls))`, using…
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: FineEnvs
Теги: lfm2, trl, openenv, harbor, agent, smoldataenvs, grpo, multi-harness
Лайков: 3 | Загрузок: 261
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.