akshayyy1/small-code-assistant

Fix chat template returning BatchEncoding instead of tensor

Add Qwen2.5-Coder-3B streaming demo on ZeroGPU

initial commit