---
title: Rombos-LLM-V2.5-Qwen-72b
canonical_url: "https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b"
md_url: "https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b.md"
repository: okwinds/Rombos-LLM-V2.5-Qwen-72b
chinese_name: "Rombos-LLM-V2.5-Qwen-72b High-performance finetuned"
last_updated: 2024-11-03
license: "Apache License 2.0"
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - qwen2
architectures:
  - Qwen2ForCausalLM
base_model:
  - Qwen/Qwen2.5-72B-Instruct
base_model_relation: finetune
parameters: 72.7B
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: 101
stars: 1
tags:
  - Chat
  - Instruct
  - fine-tuned
---

# Rombos-LLM-V2.5-Qwen-72b

> Rombos-LLM-V2.5-Qwen-72b - okwinds 在 ModelScope 开源的模型。Rombos-LLM-V2.5-Qwen-72b High-performance finetuned

okwinds/Rombos-LLM-V2.5-Qwen-72b 是 ModelScope 魔搭社区上的 72.7B 参数text-generation模型，采用 Apache License 2.0 许可，基于 Qwen/Qwen2.5-72B-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-72b
- **License**: Apache License 2.0
- **Tasks**: text-generation
- **Parameters**: 72.7B
- **Base model**: Qwen/Qwen2.5-72B-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**: 101
- **Stars**: 1
- **Last updated**: 2024-11-03

Source: https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b

---

# Rombos-LLM-V2.5-Qwen-72b 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-72b')
```
Git下载
```
#Git模型下载
git clone https://www.modelscope.cn/okwinds/Rombos-LLM-V2.5-Qwen-72b.git
```

# 模型简介

本模型转载于 huggingface 上的 [rombodawg/Rombos-LLM-V2.5-Qwen-72b](https://huggingface.co/rombodawg/Rombos-LLM-V2.5-Qwen-72b)

作者介绍到：本模型是一个基于 [Qwen/Qwen2.5-72B](https://www.modelscope.cn/models/Qwen/Qwen2.5-72B) 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/Qwen2.5-72B-Instruct](https://www.modelscope.cn/models/qwen/qwen2.5-72b-instruct) 与 [Qwen/Qwen2.5-72B](https://www.modelscope.cn/models/Qwen/Qwen2.5-72B) 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-72b](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b) </br>
	&nbsp;Base model - [Qwen/Qwen2.5-72B](https://www.modelscope.cn/models/Qwen/Qwen2.5-72B) </br>
		&nbsp; <span style="font-weight: bold;">╰</span>─  Finetuned - [Qwen/Qwen2.5-72B-Instruct](https://www.modelscope.cn/models/Qwen/Qwen2.5-72B-Instruct) </br>
		&nbsp; &nbsp; &nbsp; &nbsp; <span style="font-weight: bold;">╰</span>─  Finetuned - [本模型 okwinds/Rombos-LLM-V2.5-Qwen-72b](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b) </br>

# Benchmark

以下摘自 2024-11-3 的 Huggingface 上 [Leadboard](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) 排名数据

本模型排名第3。

截图：

<img src="https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b/resolve/master/HF-Leadboard-2024-11-3.png" />

截图中的数据：

| **Model** |**Average️**|**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|
|<span style="color:#ff6600">🔸 [***rombodawg/Rombos-LLM-V2.5-Qwen-72b***](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-72b)</span>|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|
|[rombodawg/Rombos-LLM-V2.5-Qwen-32b](https://www.modelscope.cn/models/okwinds/Rombos-LLM-V2.5-Qwen-32b)|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|
