llama.cpp for the Internet Computer.

llama_cpp_canister allows you to deploy ggerganov/llama.cpp as a Smart Contract on the Internet Computer.

• MIT open source 🧑‍💻
• Well documented 📝
• Fully QA'd via CI/CD ✅
• Easy to build, test & deploy 🚧
• Smoke testing framework using pytest 🚬

Try it out

You can try out a deployed version at https://icgpt.icpp.world/

Need help?

If you decide to use llama_cpp_canister in your ICP dApp, we want to help you.

We do NOT consider llama_cpp_canister "our IP". It is for the broad benefit of DeAI on ICP, and we hope many of you will try it out and use it.

Please join our OpenChat C++ community for any questions, discussions or feedback. ❤️

Limitations ⚠️

• Important limitation is that you can only build the canister on a Mac. (Windows & Linux is coming soon)
• The IC supports 32bit canisters only, which limits the size of deployable LLMs to ~0.5B parameters
• Everything is moving fast, so refresh your local clone frequently. ⏰
• The canister endpoint APIs are not yet 100% fixed. We might introduce breaking changes ❗❗❗

Set up

WARNING: Currently, the canister can only be build on a Mac !

• Clone the repo and it's children:

  # Clone this repo
  git clone git@github.com:onicai/llama_cpp_canister.git
  
  # Clone llama_cpp_onicai_fork, our forked version of llama.cpp
  # Into the ./src folder
  cd src
  git clone git@github.com:onicai/llama_cpp_onicai_fork.git
  
  # Initialize the submodules of the llama_cpp_onicai_fork repo
  cd llama_cpp_onicai_fork
  git submodule init
  git submodule update

• Create the file src/llama_cpp_onicai_fork/common/build-info.cpp

  make build-info-cpp-wasm
  

• Create a Python environment with dependencies installed

  ❗❗❗ Use Python 3.11 ❗❗❗

  (This is needed for binaryen.py dependency)

  # We use MiniConda
  conda create --name llama_cpp_canister python=3.11
  conda activate llama_cpp_canister
  
  # Install the python dependencies
  # From root of llama_cpp_canister repo:
  pip install -r requirements.txt

• Install dfx:

  sh -ci "$(curl -fsSL https://internetcomputer.org/install.sh)"
  
  # Configure your shell
  source "$HOME/.local/share/dfx/env"

• Build & Deploy the canister llama_cpp:

  • Compile & link to WebAssembly (wasm):

    icpp build-wasm

    Note:

    The first time you run this command, the tool-chain will be installed in ~/.icpp

    This can take a few minutes, depending on your internet speed and computer.

  • Start the local network:

    dfx start --clean

  • Deploy the wasm to a canister on the local network:

    dfx deploy
    
    # When upgrading the code in the canister, use:
    dfx deploy -m upgrade

  • Check the health endpoint of the llama_cpp canister:

    $ dfx canister call llama_cpp health
    (variant { Ok = record { status_code = 200 : nat16 } })

• Upload gguf file

  The canister is now up & running, and ready to be loaded with a gguf file. In this example we use the powerful qwen2.5-0.5b-instruct-q8_0.gguf model.

  • Download the model from huggingface: https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct-GGUF

    Store it in: models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf

  • Upload the gguf file:

    python -m scripts.upload --network local --canister llama_cpp --canister-filename models/model.gguf models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf

  • Only needed after a canister upgrade (dfx deploy -m upgrade), re-load the gguf file into Orthogonal Persisted (OP) working memory

    This step is already done by scripts.upload above, so you can skip it if you just ran that.

    After a canister upgrade, the gguf file in the canister is still there, because it is persisted in stable memory, but you need to load it into Orthogonal Persisted (working) memory, which is erased during a canister upgrade.

    dfx canister call llama_cpp load_model '(record { args = vec {"--model"; "models/model.gguf";} })'

  • Set the max_tokens for this model, to avoid it hits the IC's instruction limit

    dfx canister call llama_cpp set_max_tokens '(record { max_tokens_query = 10 : nat64; max_tokens_update = 10 : nat64 })'
    
    dfx canister call llama_cpp get_max_tokens
    

• Chat with the LLM

  • Ensure the canister is ready for Inference, with the model loaded

    dfx canister call llama_cpp ready

  • Chat with the LLM:

    Details how to use the Qwen models with llama.cpp: https://qwen.readthedocs.io/en/latest/run_locally/llama.cpp.html

    # Start a new chat - this resets the prompt-cache for this conversation
    dfx canister call llama_cpp new_chat '(record { args = vec {"--prompt-cache"; "my_cache/prompt.cache"} })'
    
    # Repeat this call until `prompt_remaining` in the response is empty. 
    # This ingest the prompt into the prompt-cache, using multiple update calls
    # Important: KEEP SENDING THE FULL PROMPT 
    dfx canister call llama_cpp run_update '(record { args = vec {"--prompt-cache"; "my_cache/prompt.cache"; "--prompt-cache-all"; "-sp"; "-p"; "<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n<|im_start|>user\ngive me a short introduction to LLMs.<|im_end|>\n<|im_start|>assistant\n"; "-n"; "512" } })' 
    ...
    
    # Once `prompt_remaining` in the response is empty, repeat this call, with an empty prompt, until `generated_eog=true`
    # Now the LLM is generating new tokens !
    dfx canister call llama_cpp run_update '(record { args = vec {"--prompt-cache"; "my_cache/prompt.cache"; "--prompt-cache-all"; "-sp"; "-p"; ""; "-n"; "512" } })'
    
    ...
    
    # Once `generated_eog` in the response is `true`, the LLM is done generating
    
    # this is the response after several update calls and it has reached eog:
    (
      variant {
        Ok = record {
          output = " level of complexity than the original text.<|im_end|>";
          conversation = "<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n<|im_start|>user\ngive me a short introduction to LLMs.<|im_end|>\n<|im_start|>assistant\nLLMs are large language models, or generative models, that can generate text based on a given input. These models are trained on a large corpus of text and are able to generate text that is similar to the input. They can be used for a wide range of applications, such as language translation, question answering, and text generation for various tasks. LLMs are often referred to as \"artificial general intelligence\" because they can generate text that is not only similar to the input but also has a higher level of complexity than the original text.<|im_end|>";
          error = "";
          status_code = 200 : nat16;
          prompt_remaining = "";
          generated_eog = true;
        }
      },
    )
    
    ########################################
    # Tip. Add this to the args vec if you #
    #      want to see how many tokens the #
    #      canister can generate before it #
    #      hits the instruction limit      #
    #                                      #
    #      ;"--print-token-count"; "1"     #
    ########################################
    

    Note: The sequence of update calls to the canister is required because the Internet Computer has a limitation on the number of computations it allows per call. At the moment, only 10 tokens can be generated per call. This sequence of update calls is equivalent to using the ggerganov/llama.cpp repo directly and running the llama-cli locally, with the command:

    ./llama-cli -m /models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf --prompt-cache prompt.cache --prompt-cache-all -sp -p "<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n<|im_start|>user\ngive me a short introduction to LLMs.<|im_end|>\n<|im_start|>assistant\n" -n 512 -fa -ngl 80 
    

    In above command, the -fa -ngl 80 arguments are useful only on GPU. We do not use them when calling the IC, because the canister has a CPU only.

• You can download the main.log file from the canister with:

  python -m scripts.download --network local --canister llama_cpp --local-filename main.log main.log
  

Smoke testing the deployed LLM

You can run a smoketest on the deployed LLM:

• Deploy Qwen2.5 model as described above

• Run the smoketests for the Qwen2.5 LLM deployed to your local IC network:

  # First test the canister functions, like 'health'
  pytest -vv test/test_canister_functions.py
  
  # Then run the inference tests
  pytest -vv test/test_qwen2.py