| --- |
| license: mit |
| language: |
| - en |
| base_model: |
| - Qwen/Qwen3-0.6B |
| tags: |
| - medical |
| - mental-health |
| --- |
| # 🧠 Qwen-0.6B Mental Health Support (Fine-Tuned) |
|
|
| **Model Repo:** `xformai/qwen-0.6b-mentalhealth-support` |
| **Base Model:** [`Qwen/Qwen-0.5B`](https://huggingface.co/Qwen/Qwen-0.5B) |
| **Task:** Empathetic Conversational AI for mental health & emotional support |
| **Fine-Tuned By:** [XformAI](https://www.linkedin.com/company/xformai) |
|
|
| --- |
|
|
| ## 🧠 What is this? |
|
|
| This is a fine-tuned version of the Qwen-0.6B language model, adapted on a curated dataset focused on mental health support and empathetic responses. The goal is to enable helpful, emotionally aware, and safe conversations around stress, anxiety, depression, and general wellness. |
|
|
| --- |
|
|
| ## 🧪 Use Cases |
|
|
| - Mental health chatbots |
| - Emotional support agents |
| - Wellness coaching prototypes |
| - Journaling assistants |
|
|
| --- |
|
|
| ## 📊 Training Details |
|
|
| - **Dataset:** Internal collection of therapy-style dialogues, emotional support threads, and curated mental health Q&A (non-clinical) |
| - **Epochs:** 3 |
| - **Batch Size:** 16 |
| - **Optimizer:** AdamW |
| - **Context Window:** 2048 |
| - **Precision:** bfloat16 |
| - **Framework:** Hugging Face Transformers + PEFT (LoRA) |
|
|
| --- |
|
|
| ## 🚨 Warnings |
|
|
| ⚠️ This model is **not a substitute for professional medical or mental health advice**. |
| It is trained to offer support-style language, not diagnosis or clinical recommendations. |
|
|
| --- |
|
|
| ## 🧠 Example Usage |
|
|
| ```python |
| from transformers import AutoModelForCausalLM, AutoTokenizer |
| |
| model = AutoModelForCausalLM.from_pretrained("xformai/qwen-0.6b-mentalhealth-support") |
| tokenizer = AutoTokenizer.from_pretrained("xformai/qwen-0.6b-mentalhealth-support") |
| |
| prompt = "I've been feeling really overwhelmed lately. Can you help?" |
| inputs = tokenizer(prompt, return_tensors="pt") |
| outputs = model.generate(**inputs, max_new_tokens=100) |
| print(tokenizer.decode(outputs[0], skip_special_tokens=True)) |
