Fine-tuned LiquidAI/LFM2.5-230M on a large synthetic reasoning dataset with preserved chain-of-thought (thinking) traces. Best eval loss: 1.6457 at step 22,000 (still improving at checkpoint time). Loss decreased from 1.866 → 1.646 over 2.88B tokens — a 12% relative reduction with clear continued downward trend at the checkpoint boundary. Uses the Liquid/LFM2 chat template with preservethinking=True. Reasoning traces from the dataset’s reasoningcontent field are mapped to the model’s native thinking field before tokenization. This is a research model fine-tuned on synthetic reasoning data. It is intended for: — Experimentation with small-model reasoning capabilities — Chain-of-thought / thinking-trace generation — Evaluation of synthetic data quality at small scale (230M) — On-device or edge reasoning model prototyping — Partial training: This checkpoint is at ~0.86 epochs (step 22,000 / 36,500 planned). The full run continues. — Small model: 230M parameters — not suitable for production deployment without further evaluation. — Synthetic data only: Trained exclusively on synthetic reasoning traces; may exhibit distribution biases from the data generation pipeline. — Limited…
Модальности:
Генерация текста
Области применения:
Логика и рассуждение Диалог / чат
Задача: Генерация текста
Автор: mkurman
Теги: lfm2, liquid, sft, reasoning, thinking, synthetic-data, conversational, en
Лайков: 3 | Загрузок: 31
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
