Instructions to use RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1 with Transformers:

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

pipe = pipeline("text-generation", model="RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1")
model = AutoModelForCausalLM.from_pretrained("RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1

SGLang

How to use RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1 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 "RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1 with Docker Model Runner:
```
docker model run hf.co/RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1
```

Qwen/Qwen3-14B W4A4 (lambda_factor=1, seed=1)

This repo contains the checkpoint exported from: /nfs/sloanlab007/projects/foundationmodelevaluation-mazumder_proj/quantization_ryan/Qwen/Qwen3-14B/W4A4/lambda_factor_1/seed_1/hf_ckpt

Quantization

weight_bits: 4
act_bits: 4
weight_quant: per_channel
act_quant: per_token

Metrics (if available)

split	dense ppl	fakequant ppl
wikitext-103 test	n/a	n/a
lambada validation	n/a	n/a

Downloads last month: 4

Safetensors

Model size

15B params

Tensor type

BF16

Model tree for RyanLucas3/ptq-Qwen-Qwen3-14B-W4A4-lf1-seed1

Base model

Qwen/Qwen3-14B-Base

Finetuned

Qwen/Qwen3-14B

Finetuned

(350)

this model