Reaction
Unsloth shares free notebook for training a Qwen3.5-4B decision model
Unsloth says Qwen3.5-4B can generate decisions instead of text locally using 8GB of VRAM.
TLDR
Unsloth shared a free notebook and guide for preparing data—states, questions and gold answers—then training and serving a Qwen3.5-4B decision model. It says the model can generate decisions instead of text locally on 8GB of VRAM.
Combined views
—
2 Sources, first seen ago
— likes— comments— saves— reposts
