File size: 1,946 Bytes
e8b8e07 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 | ---
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)) |