TinyV) is a reward system for efficient RL post-training that detects false negatives in current rule-based verifiers and provides more accurate reward signals via a small LLM during RL training. Experiments show that TinyV incurs only 6% additional computational cost while significantly increasing both RL efficiency and final model performance. — 📄 Technical Report — Including false negative analysis and theotical insights behind TinyV — 💾 Github Repo — Access the complete pipeline for more efficient RL training via TinyV — 🤗 HF Collection — Training Data, Benchmarks, and Model Artifact This model is a fine-tuned version of Qwen/Qwen3-1.7B on zhangchenxu/TinyVThinkTrainingDataQwen3_Balanced dataset. Please refer to the codebase: https://github.com/uw-nsl/TinyV for details. The following hyperparameters were used during training: — learningrate: 1e-05 — trainbatchsize: 8 — evalbatchsize: 8 — seed: 42 — distributedtype: multi-GPU — numdevices: 8 — gradientaccumulationsteps: 8 — totaltrainbatchsize: 512 — totalevalbatchsize: 64 — optimizer: Use adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments — lrschedulertype: cosine -…
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: zhangchenxu
Теги: qwen3, llama-factory, full, generated_from_trainer, conversational, text-generation-inference, endpoints_compatible
Лайков: 3 | Загрузок: 17
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
