---
title: Rombos-LLM-V2.5-Qwen-32b
canonical_url: "https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b"
md_url: "https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b.md"
repository: okwinds/Rombos-LLM-V2.5-Qwen-32b
last_updated: 2025-01-05
license: "Apache License 2.0"
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - qwen2
architectures:
  - Qwen2ForCausalLM
base_model:
  - Qwen/Qwen2.5-32B-Instruct
base_model_relation: finetune
parameters: 32.8B
tensor_type:
  - BF16
library_name:
  - safetensors
  - pytorch
frameworks:
  - Pytorch
language:
  - en
  - cn
inference_backends:
  - "deploy_task text/emb"
  - "lmdeploy 0.9.1"
  - "lmdeploy_turbomind 0.9.1"
  - "sglang 0.5.2"
  - "vllm 0.9.2"
downloads: 141
stars: 3
tags:
  - Chat
  - Instruct
  - fine-tuned
---

# Rombos-LLM-V2.5-Qwen-32b

> Rombos-LLM-V2.5-Qwen-32b - okwinds 在 ModelScope 开源的模型。Rombos-LLM-V2.5-Qwen-32b High-performance finetuned

okwinds/Rombos-LLM-V2.5-Qwen-32b 是 ModelScope 魔搭社区上的 32.8B 参数text-generation模型，采用 Apache License 2.0 许可，基于 Qwen/Qwen2.5-32B-Instruct 构建，可用 deploy_task text/emb、lmdeploy 0.9.1、lmdeploy_turbomind 0.9.1 部署。

- **Repository**: okwinds/Rombos-LLM-V2.5-Qwen-32b
- **License**: Apache License 2.0
- **Tasks**: text-generation
- **Parameters**: 32.8B
- **Base model**: Qwen/Qwen2.5-32B-Instruct
- **Inference backends**: deploy_task text/emb, lmdeploy 0.9.1, lmdeploy_turbomind 0.9.1, sglang 0.5.2, vllm 0.9.2
- **Tags**: Chat, Instruct, fine-tuned
- **Downloads**: 141
- **Stars**: 3
- **Last updated**: 2025-01-05

Source: https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b

---

# Rombos-LLM-V2.5-Qwen-32b High-performance finetuned

- ***同系合集 [Rombos-LLM-Qwen2.5](https://www.modelscope.cn/collections/Rombos-LLM-Qwen25-4dc8690ad7f541)***

## Download

SDK下载
```bash
#安装ModelScope
pip install modelscope
```
```python
#SDK模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('okwinds/Rombos-LLM-V2.5-Qwen-32b')
```
Git下载
```
#Git模型下载
git clone https://www.modelscope.cn/okwinds/Rombos-LLM-V2.5-Qwen-32b.git
```

# 模型简介

本模型转载于 huggingface 上的 [rombodawg/Rombos-LLM-V2.5-Qwen-32b](https://huggingface.co/rombodawg/Rombos-LLM-V2.5-Qwen-32b)

作者介绍到：本模型是一个基于 [Qwen/Qwen2.5-32B](https://www.modelscope.cn/models/Qwen/Qwen2.5-32B) 的 Continuous Fine-tuning 版本，并通过 [TIES-Merging](https://ar5iv.labs.arxiv.org/html/2306.01708) 方法将合并了adapter的 [Qwen/Qwen2.5-32B-Instruct](https://www.modelscope.cn/models/qwen/qwen2.5-32b-instruct) 与 [Qwen/Qwen2.5-32B](https://www.modelscope.cn/models/Qwen/Qwen2.5-32B) 进行了 TIES Merge。这使得本模型的性能优于原始的 base 模型以及 pretrain 后 fine-tune 的 Instruct 模型。

Model tree for [okwinds/Rombos-LLM-V2.5-Qwen-32b](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b) </br>
	&nbsp;Base model - [Qwen/Qwen2.5-32B](https://www.modelscope.cn/models/Qwen/Qwen2.5-32B) </br>
		&nbsp; <span style="font-weight: bold;">╰</span>─  Finetuned - [Qwen/Qwen2.5-32B-Instruct](https://www.modelscope.cn/models/Qwen/Qwen2.5-32B-Instruct) </br>
		&nbsp; &nbsp; &nbsp; &nbsp; <span style="font-weight: bold;">╰</span>─  Finetuned - [本模型 okwinds/Rombos-LLM-V2.5-Qwen-32b](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b) </br>

# Benchmark

以下摘自 2024-10-28 的 Huggingface 上 [Leadboard](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) 排名数据

本模型排名第6。

截图：

<img src="https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b/resolve/master/HF-Leadboard-2024-10-28.png" />

截图中的数据：

| **Model** |**Average**|**IFEval**|**BBH**|**MATH Lvl 5**|**GPQA**|**MUSR**|**MMLU-PRO**|
|-|-|-|-|-|-|-|-|
|dfurman/CalmeRys-78B-Orpo-v0.1 |50.78|81.63|61.92|37.92|20.02|36.37|66.8|
|MaziyarPanahi/calme-2.4-rys-78b |50.26|80.11|62.16|37.69|20.36|34.57|66.69|
|[rombodawg/Rombos-LLM-V2.5-Qwen-72b](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b) |45.39|71.55|61.27|47.58|19.8|17.32|54.83|
|dnhkng/RYS-XLarge |44.75|79.96|58.77|38.97|17.9|23.72|49.2|
|MaziyarPanahi/calme-2.1-rys-78b |44.14|81.36|59.47|36.4|19.24|19|49.38|
|<span style="color:#ff6600">🔸 [***rombodawg/Rombos-LLM-V2.5-Qwen-32b***](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b)</span>|44.1|68.27|58.26|39.12|19.57|24.73|54.62|
|MaziyarPanahi/calme-2.3-rys-78b |44.02|80.66|59.57|36.56|20.58|17|49.73|
|MaziyarPanahi/calme-2.2-rys-78b |43.92|79.86|59.27|37.92|20.92|16.83|48.73|
|MaziyarPanahi/calme-2.1-qwen2-72b |43.61|81.63|57.33|36.03|17.45|20.15|49.05|
|dnhkng/RYS-XLarge-base |43.56|79.1|58.69|34.67|17.23|22.42|49.23|
|arcee-ai/Arcee-Nova |43.5|79.07|56.74|40.48|18.01|17.22|49.47|
|MaziyarPanahi/calme-2.2-qwen2-72b |43.4|80.08|56.8|41.16|16.55|16.52|49.27|

<!-- svg class="mr-1 inline self-center flex-none" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 16 16"><path fill-rule="evenodd" clip-rule="evenodd" d="M3.9 5.34v4.07a3.68 3.68 0 0 0 3.68 3.68h1.99v-1.47H7.58c-1.21 0-2.2-1-2.2-2.21v-1.6c.63.5 1.4.76 2.2.76h1.99V7.1H7.58a2.21 2.21 0 0 1-2.16-1.76H3.9Z" fill="currentColor" fill-opacity=".5"></path><path fill-rule="evenodd" clip-rule="evenodd" d="M2.9 2.36c0-.3.25-.55.55-.55h2.43c.3 0 .55.25.55.55v2.43c0 .3-.24.55-.55.55H3.45a.55.55 0 0 1-.55-.55V2.36Zm6.67 4.23c0-.3.24-.55.55-.55h2.43c.3 0 .55.25.55.55v2.44c0 .3-.25.55-.55.55h-2.43a.55.55 0 0 1-.55-.55V6.59Zm.55 4.07c-.3 0-.55.24-.55.55v2.43c0 .3.24.55.55.55h2.43c.3 0 .55-.25.55-.55v-2.43c0-.3-.25-.55-.55-.55h-2.43Z" fill="currentColor" fill-opacity=".85"></path></svg>
Model tree for [okwinds/Rombos-LLM-V2.5-Qwen-32b](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b)
	<div class="flex flex-col space-y-0 whitespace-nowrap pl-1 text-smd text-gray-800">&nbsp; &nbsp; &nbsp;Base model - [Qwen/Qwen2.5-32B](https://www.modelscope.cn/models/Qwen/Qwen2.5-32B)</div>
		&nbsp; &nbsp; &nbsp; <svg width="15" height="28" class="text-gray-300 dark:text-gray-700" viewBox="0 0 15 18" xmlns="http://www.w3.org/2000/svg"><path fill-rule="evenodd" clip-rule="evenodd" d="M0.967742 0C0.967742 7.67991 7.21819 13.9655 15 13.9655V15C6.74768 15 0 8.3173 0 0H0.967742Z" fill="currentColor"></path></svg>
		<span class="mr-auto flex items-center font-semibold"><svg class="mr-1 inline self-center flex-none text-gray-400" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 17 17"><path opacity="0.25" d="M12.1072 5.27442L7.39274 8.00113L2.67828 5.27442C2.77506 5.10687 2.9131 4.96686 3.07927 4.86771L6.8199 2.70811C7.17506 2.50762 7.61042 2.50762 7.96558 2.70811L11.7062 4.86771C11.8723 4.96681 12.0104 5.10715 12.1072 5.27442Z" fill="currentColor"></path><path opacity="0.5" d="M7.3928 8.00112V13.4431C7.21219 13.4428 7.03398 13.4017 6.87152 13.3228L3.09652 11.1403C2.92264 11.0397 2.77822 10.8952 2.6777 10.7213C2.57717 10.5474 2.52406 10.3501 2.52368 10.1493V5.85298C2.52463 5.65003 2.5779 5.45076 2.67835 5.27441L7.3928 8.00112Z" fill="currentColor"></path><path fill-rule="evenodd" clip-rule="evenodd" d="M12.2621 7.17681V5.85298C12.261 5.6502 12.2077 5.45085 12.1075 5.27441L7.39301 8.00112V13.4431C7.56715 13.4419 7.73957 13.403 7.8971 13.3285L9.44011 12.4381C9.39138 12.3915 9.34927 12.338 9.31521 12.2791C9.25238 12.1704 9.21919 12.0471 9.21895 11.9215V9.23635C9.21955 9.10957 9.25281 8.98509 9.31553 8.87491L9.31524 8.87474C9.37573 8.77002 9.462 8.68251 9.56586 8.62054L11.9037 7.27079C12.0148 7.20808 12.1385 7.17675 12.2621 7.17681Z" fill="currentColor"></path><path opacity="0.25" d="M15.2083 8.87475L12.2617 10.5789L9.31519 8.87475C9.37567 8.77002 9.46195 8.68252 9.5658 8.62055L11.9037 7.2708C12.1257 7.14549 12.3978 7.14549 12.6197 7.2708L14.9576 8.62055C15.0615 8.68249 15.1477 8.7702 15.2083 8.87475Z" fill="currentColor"></path><path opacity="0.5" d="M12.2621 10.5789V13.9801C12.1492 13.9799 12.0378 13.9542 11.9363 13.9049L9.5769 12.5409C9.46822 12.478 9.37796 12.3877 9.31513 12.279C9.2523 12.1703 9.21911 12.047 9.21887 11.9215V9.2363C9.21946 9.10946 9.25276 8.98491 9.31554 8.87469L12.2621 10.5789Z" fill="currentColor"></path><path d="M15.3052 9.2363V11.9215C15.305 12.047 15.2718 12.1703 15.209 12.279C15.1461 12.3877 15.0559 12.478 14.9472 12.5409L12.5771 13.9085C12.4786 13.9551 12.3709 13.9794 12.262 13.9801V10.5789L15.2086 8.87469C15.2712 8.98497 15.3045 9.10956 15.3052 9.2363Z" fill="currentColor"></path></svg> Finetuned - [Qwen/Qwen2.5-32B-Instruct](https://www.modelscope.cn/models/Qwen/Qwen2.5-32B-Instruct)</span>
		<div class="flex flex-col space-y-0 whitespace-nowrap pl-1 text-smd text-gray-800">
		&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; <svg width="15" height="28" class="text-gray-300 dark:text-gray-700" viewBox="0 0 15 18" xmlns="http://www.w3.org/2000/svg"><path fill-rule="evenodd" clip-rule="evenodd" d="M0.967742 0C0.967742 7.67991 7.21819 13.9655 15 13.9655V15C6.74768 15 0 8.3173 0 0H0.967742Z" fill="currentColor"></path></svg>
		<span class="mr-auto flex items-center font-semibold"><svg class="mr-1 inline self-center flex-none text-gray-400" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 17 17"><path opacity="0.25" d="M12.1072 5.27442L7.39274 8.00113L2.67828 5.27442C2.77506 5.10687 2.9131 4.96686 3.07927 4.86771L6.8199 2.70811C7.17506 2.50762 7.61042 2.50762 7.96558 2.70811L11.7062 4.86771C11.8723 4.96681 12.0104 5.10715 12.1072 5.27442Z" fill="currentColor"></path><path opacity="0.5" d="M7.3928 8.00112V13.4431C7.21219 13.4428 7.03398 13.4017 6.87152 13.3228L3.09652 11.1403C2.92264 11.0397 2.77822 10.8952 2.6777 10.7213C2.57717 10.5474 2.52406 10.3501 2.52368 10.1493V5.85298C2.52463 5.65003 2.5779 5.45076 2.67835 5.27441L7.3928 8.00112Z" fill="currentColor"></path><path fill-rule="evenodd" clip-rule="evenodd" d="M12.2621 7.17681V5.85298C12.261 5.6502 12.2077 5.45085 12.1075 5.27441L7.39301 8.00112V13.4431C7.56715 13.4419 7.73957 13.403 7.8971 13.3285L9.44011 12.4381C9.39138 12.3915 9.34927 12.338 9.31521 12.2791C9.25238 12.1704 9.21919 12.0471 9.21895 11.9215V9.23635C9.21955 9.10957 9.25281 8.98509 9.31553 8.87491L9.31524 8.87474C9.37573 8.77002 9.462 8.68251 9.56586 8.62054L11.9037 7.27079C12.0148 7.20808 12.1385 7.17675 12.2621 7.17681Z" fill="currentColor"></path><path opacity="0.25" d="M15.2083 8.87475L12.2617 10.5789L9.31519 8.87475C9.37567 8.77002 9.46195 8.68252 9.5658 8.62055L11.9037 7.2708C12.1257 7.14549 12.3978 7.14549 12.6197 7.2708L14.9576 8.62055C15.0615 8.68249 15.1477 8.7702 15.2083 8.87475Z" fill="currentColor"></path><path opacity="0.5" d="M12.2621 10.5789V13.9801C12.1492 13.9799 12.0378 13.9542 11.9363 13.9049L9.5769 12.5409C9.46822 12.478 9.37796 12.3877 9.31513 12.279C9.2523 12.1703 9.21911 12.047 9.21887 11.9215V9.2363C9.21946 9.10946 9.25276 8.98491 9.31554 8.87469L12.2621 10.5789Z" fill="currentColor"></path><path d="M15.3052 9.2363V11.9215C15.305 12.047 15.2718 12.1703 15.209 12.279C15.1461 12.3877 15.0559 12.478 14.9472 12.5409L12.5771 13.9085C12.4786 13.9551 12.3709 13.9794 12.262 13.9801V10.5789L15.2086 8.87469C15.2712 8.98497 15.3045 9.10956 15.3052 9.2363Z" fill="currentColor"></path></svg> Finetuned - [本模型](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b)</span></div//-->

<br/>
_______________________________

#### _作者在此 👇🏻 扫一扫_

<img src="https://www.modelscope.cn/models/okwinds/GPT-2/resolve/master/qrcode_for_jcl_258.jpg" />
