README.md 3.71 KB
Newer Older
jerrrrry's avatar
jerrrrry committed
1
<!-- EasyStart v0.1 完整使用手册(单文件版) -->
jerrrrry's avatar
jerrrrry committed
2
<p align="center">
jerrrrry's avatar
jerrrrry committed
3
4
5
6
7
8
9
10
11
12
13
14
15
16
<pre style="font-family: monospace; text-align: center; border: none;">
     ╭─────────────────────────╮
     │  ⚡  EasyStart v0.1  ⚡  │
     ╰─────────────────────────╯
            ╱╲
           ╱  ╲
          ╱    ╲      ROCm & DCU
         ╱  ❖   ╲
        ╱────────╲
       ╱  ▖    ▗  ╲
      ╱────────────╲
     ╱  ▖  ▖  ▗  ▗  ╲
    ╱────────────────╲
</pre>
jerrrrry's avatar
jerrrrry committed
17
</p>
jerrrrry's avatar
jerrrrry committed
18

jerrrrry's avatar
jerrrrry committed
19
20
21
<h1 align="center">
  EasyStart v0.1 —— 一键启动,零门槛大模型测试
</h1>
jerrrrry's avatar
jerrrrry committed
22

jerrrrry's avatar
jerrrrry committed
23
<p align="center">
jerrrrry's avatar
jerrrrry committed
24
25
26
  <a href="#scene1">环境测试</a>
  <a href="#scene2">测试+下载+推理</a>
  <a href="#scene3">批量本地推理</a>
jerrrrry's avatar
jerrrrry committed
27
</p>
jerrrrry's avatar
jerrrrry committed
28

jerrrrry's avatar
jerrrrry committed
29
---
jerrrrry's avatar
jerrrrry committed
30

jerrrrry's avatar
jerrrrry committed
31
32
> **一句话总结**  
> 无论做交付、做评测还是做批量实验,只要一条命令,环境、模型、推理全搞定。
jerrrrry's avatar
jerrrrry committed
33

jerrrrry's avatar
jerrrrry committed
34
---
jerrrrry's avatar
jerrrrry committed
35

jerrrrry's avatar
jerrrrry committed
36
## 🚀 快速开始
jerrrrry's avatar
jerrrrry committed
37

jerrrrry's avatar
jerrrrry committed
38
39
| 场景 | 一键指令 |
|------|-----------|
jerrrrry's avatar
jerrrrry committed
40
41
42
| 1️⃣ 纯环境测试 | `git clone http://developer.sourcefind.cn/codes/jerrrrry/easystart_v0.1.git && cd easystart_v0.1/1_env_check && bash start.sh` |
| 2️⃣ 环境测试 + 模型下载 + 大模型推理 | `git clone http://developer.sourcefind.cn/codes/jerrrrry/easystart_v0.1.git && cd "easystart_v0.1/2_env_check&model_download&llm_inference" && bash start.sh` |
| 3️⃣ 环境测试 + 批量本地模型推理 | `git clone http://developer.sourcefind.cn/codes/jerrrrry/easystart_v0.1.git && cd "easystart_v0.1/3_env_check&batches_llm_inference" && bash start.sh` |
jerrrrry's avatar
jerrrrry committed
43

jerrrrry's avatar
jerrrrry committed
44
---
jerrrrry's avatar
jerrrrry committed
45

jerrrrry's avatar
jerrrrry committed
46
47
<a name="scene1"></a>
## 📦 1️⃣ 环境测试(`1_env_check`)
jerrrrry's avatar
jerrrrry committed
48
49
50
51
52
53
54
- ✅ ROCm 带宽测试  
- ✅ 4/8 卡 RCCL 带宽  
- ✅ DCU 环境检查(贵哥发版)  
- ✅ ACS 监控  
- ✅ CPU & DCU 状态  
- ✅ 存储 & 内存  
- ✅ 网络连通性  
jerrrrry's avatar
jerrrrry committed
55

jerrrrry's avatar
jerrrrry committed
56
📁 结果输出:`./outputs/env_check_outputs`
jerrrrry's avatar
jerrrrry committed
57

jerrrrry's avatar
jerrrrry committed
58
59
60
<p align="center">
  <img src="images/1.png" width="600"/>
</p>
jerrrrry's avatar
jerrrrry committed
61

jerrrrry's avatar
jerrrrry committed
62
---
jerrrrry's avatar
jerrrrry committed
63

jerrrrry's avatar
jerrrrry committed
64
65
<a name="scene2"></a>
## 📦 2️⃣ 环境测试 + 模型下载 + 推理(`2_env_check&model_download&llm_inference`)
jerrrrry's avatar
jerrrrry committed
66

jerrrrry's avatar
jerrrrry committed
67
### ① 填写待测模型  
jerrrrry's avatar
jerrrrry committed
68
`download-list.cfg` 中按以下格式添加模型:
jerrrrry's avatar
jerrrrry committed
69
模型ID;本地保存路径(模型ID对应modelscope的模型ID)
jerrrrry's avatar
jerrrrry committed
70
71
72
73
> 可一次填写多个,支持批量下载与测试。  
<p align="center">
  <img src="images/3.png" width="400"/>
</p>
jerrrrry's avatar
jerrrrry committed
74

jerrrrry's avatar
jerrrrry committed
75
76
77
78
79
80
81
82
83
### ② 配置推理参数  
编辑 `model_to_test.cfg`,按指定格式填入推理参数。  
<p align="center">
  <img src="images/4.png" width="400"/>
</p>

### ③ 运行脚本
```bash
bash start.sh
jerrrrry's avatar
jerrrrry committed
84
```
jerrrrry's avatar
jerrrrry committed
85
86
87



jerrrrry's avatar
jerrrrry committed
88

jerrrrry's avatar
jerrrrry committed
89
90
91
92
93
### ④ 结果查看

环境报告:./outputs/env_check_outputs
推理结果:./outputs/inference_outputs
下载模型:./outputs/models
jerrrrry's avatar
jerrrrry committed
94

jerrrrry's avatar
jerrrrry committed
95
96
97
98
99
<p align="center">
  <img src="images/5.png" width="600"/>
  <img src="images/6.png" width="600"/>
  <img src="images/7.png" width="600"/>
</p>
jerrrrry's avatar
jerrrrry committed
100
101
102
103
104
105




<a name="scene3"></a>
## 📦 3️⃣ 环境测试 + 批量本地模型推理(3_env_check&batches_llm_inference)
jerrrrry's avatar
jerrrrry committed
106
107
108

### ① 挂载本地模型

jerrrrry's avatar
jerrrrry committed
109
110
只需在start.sh中挂载本地大模型到docker里

jerrrrry's avatar
jerrrrry committed
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
-v /your/local/model/path:/workspace/models


### ② 配置推理参数

编辑同目录下的 model_to_test.cfg,按指定格式填入测试参数。

### ③ 运行脚本
```bash
bash start.sh
```

### ④ 结果查看

推理结果统一输出到 ./outputs/inference_outputs

<p align="center">
  <img src="images/8.png" width="600"/>
</p>


## 📝 小贴士
所有脚本均基于 Docker,确保已安装 Docker & ROCm 环境。
建议首次运行前执行场景 1,确认环境无虞。
遇到任何问题,欢迎提 Issue。

<p align="center">
jerrrrry's avatar
jerrrrry committed
138
  ❤️
jerrrrry's avatar
jerrrrry committed
139
140
</p>
```