Instructions to use meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024 with Transformers:

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

pipe = pipeline("text-generation", model="meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024")
model = AutoModelForCausalLM.from_pretrained("meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024", 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024

SGLang

How to use meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024 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 "meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024" \
    --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": "meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024",
		"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 "meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024" \
        --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": "meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024 with Docker Model Runner:
```
docker model run hf.co/meteorain/Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024
```

Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024 / README.md

meteorain

Add llm-compressor gptq n256 s1024

0fe33a2 verified about 2 months ago

preview code

Raw

History Blame Contribute Delete

539 Bytes

metadata

library_name: transformers
base_model:
  - Qwen/Qwen3-4B-Thinking-2507
base_model_relation: quantized
pipeline_tag: text-generation
tags:
  - Qwen3
  - GPTQ
  - W4A16
  - llm-compressor

Qwen3-4B-Thinking-2507-llmc-gptq-w4a16-g128-n256-s1024

Base model: Qwen/Qwen3-4B-Thinking-2507

Quantized with llm-compressor.

method: gptq
weight format: W4A16
group size: 128
calibration dataset: nvidia/Llama-Nemotron-Post-Training-Dataset
calibration split: math
calibration samples: 256
max sequence length: 1024