• Recent
  • Wiki
  • Github
  • 百度网盘
  • Onedrive
  • Official
  • Shop
  • Register
  • Login
  • Register
  • Login
  • Search
  • Recent
  • Wiki
  • Github
  • 百度网盘
  • Onedrive
  • Official
  • Shop

Deploying Deepseek R7 model on coolpi

AI Algorithm
1
1
335
Loading More Posts
  • Oldest to Newest
  • Newest to Oldest
  • Most Votes
Reply
  • Reply as topic
Log in to reply
This topic has been deleted. Only users with topic management privileges can see it.
  • G
    george last edited by george Feb 14, 2025, 12:38 AM Feb 11, 2025, 2:24 AM

    The R7 model is suitable for cp4b (16G) cm5-evb (32G) cm5-notebook (32G) machines, and low memory capacity machines are only suitable for the R1.5 model. The operating system defaults to Ubuntu 24.04.

    • Update the system loader to the latest version
      Download the latest loader file from Baidu Netdisk or OneDrive
      627c03ed-d3aa-4d95-ba7a-819378159e52-image.png
      The machine enters Maskrom and then upgrades the loader according to the configuration shown in the following figure.
      673b0e95-ef68-4f93-90cd-732c093651de-image.png
    • Upgrade to the latest Ubuntu 24.04 image
      Download the latest Ubuntu 24.04 image from Baidu Netdisk or OneDrive and upgrade. Currently, CP4B/CM5-EVB/CM5-Notebook uses a unified image file.
      ef6751f9-ee96-422b-8ba3-f0f0cc73557e-image.png
      The machine enters UMS mode (run the command "sudo reboot UMS" in a shell window or press the power button multiple times after booting), and upgrade using Etcher.
      Attention: The downloaded image needs to be decompressed before it can be used.
      33130c91-11c8-4e5c-b74a-8fa230b67572-image.png
    • Download Model
      Obtain the converted rknn model from the following website. Deepseek can choose the 1.5b/7b model. Of course, if you are interested, you can convert it yourself. Tools and documentation will be provided later.
      https://huggingface.co/c01zaut
      e73c6548-78b3-4a77-9593-1b0d2495d540-image.png
    • Synchronize RKLMM project
    git clone https://github.com/coolpi-george/rknn-llm.git
    cd rknn-llm/examples/DeepSeek-R1-Distill-Qwen-1.5B_Demo/deploy/
    ./build-linux.sh 
    

    Compiled successfully as shown in the following figure

    tmjs@tmjs:~/share/coolpi-george/rknn-llm/examples/DeepSeek-R1-Distill-Qwen-1.5B_Demo/deploy$ ./build-linux.sh
    -- Configuring done (0.0s)
    -- Generating done (0.0s)
    -- Build files have been written to: /home/tmjs/share/coolpi-george/rknn-llm/examples/DeepSeek-R1-Distill-Qwen-1.5B_Demo/deploy/build/build_linux_aarch64_Release
    [100%] Built target llm_demo
    [100%] Built target llm_demo
    Install the project...
    -- Install configuration: "Release"
    -- Up-to-date: /home/tmjs/share/coolpi-george/rknn-llm/examples/DeepSeek-R1-Distill-Qwen-1.5B_Demo/deploy/install/demo_Linux_aarch64/./llm_demo
    -- Up-to-date: /home/tmjs/share/coolpi-george/rknn-llm/examples/DeepSeek-R1-Distill-Qwen-1.5B_Demo/deploy/install/demo_Linux_aarch64/lib/librkllmrt.so
    
    sudo cp install/demo_Linux_aarch64/lib/librkllmrt.so /usr/lib/aarch64-linux-gnu/
    
    • Load the model and run it
    tmjs@tmjs:~/share/coolpi-george/rknn-llm/examples/DeepSeek-R1-Distill-Qwen-1.5B_Demo/deploy/install/demo_Linux_aarch64$ ./llm_demo /home/tmjs/share/DeepSeek-R1-Distill-Qwen-1.5B_RKLLM/deepseek-r1-1.5B-rkllm1.1.4.rkllm 100000 100000
    rkllm init start
    I rkllm: rkllm-runtime version: 1.1.4, rknpu driver version: 0.9.7, platform: RK3588
    
    rkllm init success
    
    **********************可输入以下问题对应序号获取回答/或自定义输入********************
    
    [0] 现有一笼子,里面有鸡和兔子若干只,数一数,共有头14个,腿38条,求鸡和兔子各有多少只?
    [1] 有28位小朋友排成一行,从左边开始数第10位是学豆,从右边开始数他是第几位?
    
    *************************************************************************
    
    
    user: 你好
    robot: <think>
    
    </think>
    
    你好!很高兴见到你,有什么我可以帮忙的吗?无论是学习、工作还是生活中的问题,都可以告诉我哦!
    
    
    user: 帮我写一首赞美春天的歌曲
    robot: <think>
    嗯,用户让我帮他写一首赞美春天的歌曲。首先,我得考虑用户的需求是什么。他可能是个学生,老师布置了这样的作业,或者他自己想表达对春天的喜爱。也有可能是喜欢音乐创作的人,想要一首有诗意、充满生机的作品。
    
    接下来,我要分析用户可能没有明确说出来的深层需求。也许他希望这首诗不仅要有春天的意象,还要有情感上的共鸣,比如对自然美景的赞美,或者是对未来的美好期许。另外,用户可能还希望这首诗能够传达一种宁静而美好的氛围,适合在春天的早晨或午后听。
    
    然后,我得考虑如何将春天的元素融入到歌曲里。春天通常与花开、绿草、鸟鸣、风声等相关联。这些元素可以用来营造出春天的氛围,同时让歌曲有节奏感和韵律感。比如,用轻快的鼓点来表现风的声音,或者用轻柔的旋律来描绘花开的美丽。
    
    我还得考虑诗的结构和押韵方式。可能采用四句一节,每句七个字,这样比较符合中文诗歌的韵律。同时,每一句都要有独特的意象,避免重复,让整首诗更有层次感。
    
    另外,用户可能希望这首诗不仅仅是描述春天,还能传达一种积极向上的情绪。因此,在用词和节奏上要显得温暖而充满希望。比如,使用一些积极向上的词汇,如“绿”、“欢”、“暖”等,来增强整体的感染力。
    
    最后,我需要确保整首诗流畅自然,情感真挚。可能还需要加入一些比喻或拟人手法,让春天的景象更加生动形象。例如,把花朵比作小舞者,或者把风的声音比作春天的旋律。
    
    总结一下,用户的需求是写一首赞美春天的歌曲,我需要结合春天的元素,使用富有节奏感和韵律的结构,传达出宁静而美好的氛围,并且情感真挚。这样既满足了用户的创作需求,也能让这首诗打动人心。
    </think>
    
    ## 《春日》
    
    花儿在枝头摇曳
    轻盈的裙摆上
    藏着春天的旋律
    风儿轻轻吹过
    带来泥土的气息
    我听见了花开的声音
    在耳畔流淌
    每一朵花都是我的家
    它们是春天的歌谣
    在记忆里
    等待着
    我听见了
    春天的节奏
    在心上
    它是一首永不停歇的诗
    
    1 Reply Last reply Reply Quote 0
    • 1 / 1
    1 out of 1
    • First post
      1/1
      Last post