Instructions to use GMLHUHE/PsyLLM-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GMLHUHE/PsyLLM-4B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="GMLHUHE/PsyLLM-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("GMLHUHE/PsyLLM-4B")
model = AutoModelForCausalLM.from_pretrained("GMLHUHE/PsyLLM-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use GMLHUHE/PsyLLM-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "GMLHUHE/PsyLLM-4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GMLHUHE/PsyLLM-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/GMLHUHE/PsyLLM-4B

SGLang

How to use GMLHUHE/PsyLLM-4B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "GMLHUHE/PsyLLM-4B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GMLHUHE/PsyLLM-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "GMLHUHE/PsyLLM-4B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GMLHUHE/PsyLLM-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use GMLHUHE/PsyLLM-4B with Docker Model Runner:
```
docker model run hf.co/GMLHUHE/PsyLLM-4B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

PsyLLM is a large language model designed for psychological counseling and mental health dialogue generation. It integrates diagnostic reasoning and therapeutic reasoning, following established frameworks such as DSM/ICD, and incorporates diverse therapeutic approaches including CBT, ACT, and psychodynamic therapy.

The model is trained on the OpenR1-Psy dataset (arXiv:2505.15715),
featuring multi-turn counseling dialogues with explicit reasoning traces that support clinically informed, empathetic, and interpretable AI-assisted therapy.

The training process is implemented based on the open-source framework LLaMA-Factory. If you find this project helpful, feel free to ⭐ it! PsyLLM

推理示例代码

from transformers import AutoModelForCausalLM, AutoTokenizer

model_path = "GMLHUHE/PsyLLM-4B"

# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(
    model_path,
    torch_dtype="auto",
    device_map="auto"
)

# prepare the model input
prompt = "I have participated in big group sessions before where I was left to find my own safe place, but it hasn't worked for me."
messages = [
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
    enable_thinking=True 
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

# conduct text completion
generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()

# parsing thinking content
try:
    index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
    index = 0

thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")

print("PsyLLM thinking content:", thinking_content)
print("PsyLLM content:", content)

📄 Citation

If you use this dataset, please cite:

@article{hu2025beyond,
  title={Beyond Empathy: Integrating Diagnostic and Therapeutic Reasoning with Large Language Models for Mental Health Counseling},
  author={Hu, He and Zhou, Yucheng and Si, Juzheng and Wang, Qianning and Zhang, Hengheng and Ren, Fuji and Ma, Fei and Cui, Laizhong},
  journal={arXiv preprint arXiv:2505.15715},
  year={2025}
}