Instructions to use TCMLLM/Lingdan-4B-R1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TCMLLM/Lingdan-4B-R1 with Transformers:

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

pipe = pipeline("text-generation", model="TCMLLM/Lingdan-4B-R1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("TCMLLM/Lingdan-4B-R1")
model = AutoModelForCausalLM.from_pretrained("TCMLLM/Lingdan-4B-R1", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use TCMLLM/Lingdan-4B-R1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TCMLLM/Lingdan-4B-R1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TCMLLM/Lingdan-4B-R1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/TCMLLM/Lingdan-4B-R1

SGLang

How to use TCMLLM/Lingdan-4B-R1 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 "TCMLLM/Lingdan-4B-R1" \
    --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": "TCMLLM/Lingdan-4B-R1",
		"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 "TCMLLM/Lingdan-4B-R1" \
        --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": "TCMLLM/Lingdan-4B-R1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use TCMLLM/Lingdan-4B-R1 with Docker Model Runner:
```
docker model run hf.co/TCMLLM/Lingdan-4B-R1
```

TCMLLM commited on 6 days ago

Commit

ebf05aa

verified ·

1 Parent(s): c848ca3

Document cumulative CPT, SFT and RL training stages

Browse files

Files changed (1) hide show

README.md +12 -2

README.md CHANGED Viewed

@@ -39,17 +39,27 @@ tags:
 ## Model Details
-**Lingdan-4B-R1** is a Qwen3-based TCM model in the [Lingdan-V2 family](https://github.com/TCMAI-BJTU/Lingdan-V2). CPT → SFT → prescription-centered GRPO on 24,212 samples.
 | Field | Value |
 | --- | --- |
 | Parameter scale | 4B |
 | Architecture family | Qwen3 causal language model |
-| Training stage | Prescription-centered GRPO |
 | Starting checkpoint / backbone | Lingdan-4B-SFT |
 | Intended use | Structured TCM clinical reasoning and prescription-generation research |
 | Platform model ID | `TCMLLM/Lingdan-4B-R1` |
 ## Checkpoint Evaluation
 Paper-reported results for **Lingdan-4B-R1**. Full: LingLan (25,620 items); Hard: LingLan-Hard (5,200 items). Overall Score averages 16 primary metrics, excluding dose MAE. All scores use a 0–100 scale except MAE (lower is better). The paper uses zero-shot Chinese prompts, temperature 0.6, and at most 8,192 generated tokens; reproduction also requires the same data and scoring pipeline.

 ## Model Details
+**Lingdan-4B-R1** is a Qwen3-based TCM model in the [Lingdan-V2 family](https://github.com/TCMAI-BJTU/Lingdan-V2).
 | Field | Value |
 | --- | --- |
 | Parameter scale | 4B |
 | Architecture family | Qwen3 causal language model |
+| Training stage | Reinforcement learning (RL; prescription-centered GRPO) |
 | Starting checkpoint / backbone | Lingdan-4B-SFT |
 | Intended use | Structured TCM clinical reasoning and prescription-generation research |
 | Platform model ID | `TCMLLM/Lingdan-4B-R1` |
+### Training History
+This checkpoint includes the following training stages, in order:
+| Stage | Objective | Training data |
+| --- | --- | --- |
+| CPT — continued pretraining | TCM-domain knowledge acquisition | 16.62B tokens |
+| SFT — supervised fine-tuning | Reasoning-oriented instruction following | 7,729,982 samples / 2.08B tokens |
+| RL — prescription-centered GRPO | Prescription composition alignment | 24,212 samples |
 ## Checkpoint Evaluation
 Paper-reported results for **Lingdan-4B-R1**. Full: LingLan (25,620 items); Hard: LingLan-Hard (5,200 items). Overall Score averages 16 primary metrics, excluding dose MAE. All scores use a 0–100 scale except MAE (lower is better). The paper uses zero-shot Chinese prompts, temperature 0.6, and at most 8,192 generated tokens; reproduction also requires the same data and scoring pipeline.