Instructions to use SETT-Centre/chatty_mapper with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SETT-Centre/chatty_mapper with Transformers:

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

pipe = pipeline("text-generation", model="SETT-Centre/chatty_mapper")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("SETT-Centre/chatty_mapper")
model = AutoModelForCausalLM.from_pretrained("SETT-Centre/chatty_mapper")
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 SETT-Centre/chatty_mapper with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/SETT-Centre/chatty_mapper

SGLang

How to use SETT-Centre/chatty_mapper 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 "SETT-Centre/chatty_mapper" \
    --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": "SETT-Centre/chatty_mapper",
		"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 "SETT-Centre/chatty_mapper" \
        --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": "SETT-Centre/chatty_mapper",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use SETT-Centre/chatty_mapper with Docker Model Runner:
```
docker model run hf.co/SETT-Centre/chatty_mapper
```

A First Attempt at Getting LLMs to map SNOMED codes

This model is really poor and represents a first stab at finetuning Mixtral to map SNOMED codes. It doesn't really work and I would recommend using a newer model as this one is way out of date now.

Model description

This is a text generation model for SNOMED-CT. As it is text-generation, it is prone to hallucination and should not be used for any kind of production purpose but it was fun to build. It is based on Mixtral7b and was fine-tuned on a part of the SNOMED-CT corpus then tested against a gold-standard.

How to use

Provide code snippets on how to use your model.

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "MattStammers/chatty_mapper"
model = AutoModelForCausalLM.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)

# Your example here
Model Performance
Accuracy: 0.0
Precision: 0.0
Recall: 0.0
Example DataFrame head:                   ParameterName  SNOMEDCode  \
0                   *Heart rate   364075005   
1  Peripheral oxygen saturation   431314004   
2        Mean arterial pressure  1285244000   
3     *Diastolic blood pressure   271650006   
4      *Systolic blood pressure   271649006   

                              ExtractedSNOMEDNumbers  CorrectPrediction  
0                                            3222222              False  
1  4222222000000000000000000000000000000000000000...              False  
2                                                NaN              False  
3                                                NaN              False  
4                                                NaN              False  

Limitations and bias
It is prone to wandering and certainly not medical-grade.

Acknowledgments
Thanks to the Mixtral AI team for creating the base model.

Save the model card in the model directory with open(f"models/chatty_mapper/README.md", "w") as f: f.write(model_card_content)

Use Hugging Face's Repository class for Git operations repo = Repository(local_dir=model_save_path, clone_from=repo_url) repo.git_add() repo.git_commit("Initial model upload with model card and metrics") repo.git_push()

print(f"Model, model card, and metrics successfully pushed to: https://huggingface.co/MattStammers/chatty_mapper")

Downloads last month: 6

Safetensors

Model size

7B params

Tensor type

F32