Vortex
Collection
ModelCloud optimized and validated quants that pass/meet strict quality assurance on multiple benchmarks. No one quantize • 24 items • Updated • 10
Text Generation
Safetensors
English
qwen3
gptqmodel
modelcloud
chat
marin
instruct
int4
gptq
4bit
w4a16
4-bit precision
How to use from
vLLMUse Docker
docker model run hf.co/ModelCloud/Marin-32B-Base-GPTQMODEL-W4A16Quick Links
This 4bit W4A16 model has been quantized by @Qubitum at ModelCloud using GPT-QModel.
SPDX-License-Identifier: NonCommercial-ModelCloud
Copyright (c) 2024–2025 ModelCloud.ai
Released for non-profit use only.
Commercial use requires permission from [@qubitium](https://x.com/qubitium).
| Metric | Marin |
|--------------------------------|----------|
| arc_challenge :: acc,none | 0.5725 |
| arc_challenge :: acc_norm,none | 0.6007 |
| mmlu_stem :: acc,none | 0.667 |
- Downloads last month
- 7
Model tree for ModelCloud/Marin-32B-Base-GPTQMODEL-W4A16
Base model
marin-community/marin-32b-base
Install from pip and serve model
# Install vLLM from pip: pip install vllm# Start the vLLM server: vllm serve "ModelCloud/Marin-32B-Base-GPTQMODEL-W4A16"# Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ModelCloud/Marin-32B-Base-GPTQMODEL-W4A16", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'