Instructions to use properexit/ArgParser-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use properexit/ArgParser-v1 with Transformers:

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

pipe = pipeline("text-generation", model="properexit/ArgParser-v1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("properexit/ArgParser-v1")
model = AutoModelForCausalLM.from_pretrained("properexit/ArgParser-v1")
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 Settings

vLLM

How to use properexit/ArgParser-v1 with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/properexit/ArgParser-v1

SGLang

How to use properexit/ArgParser-v1 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 "properexit/ArgParser-v1" \
    --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": "properexit/ArgParser-v1",
		"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 "properexit/ArgParser-v1" \
        --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": "properexit/ArgParser-v1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use properexit/ArgParser-v1 with Docker Model Runner:
```
docker model run hf.co/properexit/ArgParser-v1
```

ArgParser-v1

Baseline for the ArgParser series. Full fine-tune of Qwen-0.5B on four argument-mining corpora (AbstRCT, Microtext, CDCP, PERSPECTRUM), 1,494 records total, 3 epochs, fp16, Adafactor. About 1.5 hours on a GTX 1080 Ti.

Held-out component-F1 averaged across the four domains: 0.108. Best on CDCP claim extraction (0.501). Worst on PERSPECTRUM (91% empty rate — the debate-text format defeats extractive parsing here).

This is the smallest useful reference point. Kept up mostly for reproducibility of the ablation series. If you want to actually use one of these, use ArgParser-v4 — same repo family, gets a real Phase 1 integration F1 of 0.217 versus this baseline's near-zero usefulness on that task.