EmoLLM

Why we included this project

EmoLLM is a family of instruction-tuned models for Chinese-language mental health support, built for supportive, counseling-style conversations. The repository keeps the whole pipeline in one place: fine-tuning configs for InternLM, Qwen, Llama, GLM, and other base models, plus curated training datasets, evaluation scripts, deployment guides, and a RAG path for grounding responses in reference material. Most chatbot projects stop at a model download; here you get the steps to turn a base model into a working counseling assistant and keep improving it. Researchers studying empathetic dialogue and product teams that want to ship a psychological support assistant will both find something practical to build on.