example-multi-thread.html

<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="utf-8">
    <meta name="viewport" content="width=device-width, initial-scale=1">
    <title>llama-cpp-wasm multithreading</title>
    <link rel="icon" type="image/png" href="favicon.png" />
    
    <!-- picocss -->
    <link
      rel="stylesheet"
      href="https://cdn.jsdelivr.net/npm/@picocss/pico@2/css/pico.min.css"
    />
  </head>

  <body>
    <header class="container">
      <hgroup>
        <h1><a href="/">llama-cpp-wasm</a> &nbsp; &#128007; <mark>multithreading</mark> wasm32 </h1>
        <br />
        
        <p> WebAssembly (Wasm) Build and Bindings for <a href="https://github.com/ggerganov/llama.cpp" target="_blank">llama.cpp</a>. </p>
        <br />
        
        <p> This demonstration enables you to run LLM models directly in your browser utilizing JavaScript, WebAssembly, and llama.cpp. </p>
        <br />

        <p> Repository: <a href="https://github.com/tangledgroup/llama-cpp-wasm"> https://github.com/tangledgroup/llama-cpp-wasm </a></p>
        <br />

        <p> When you click <b>Run</b>, model will be first downloaded and cached in browser. </p>
      </hgroup>
    </header>

    <main class="container">
      <section>
        <h2> Demo </h2>
        
        <label> Model: </label>

        <select id="model" name="model" aria-label="Select model" required>
          <!-- <option value="https://huggingface.co/Qwen/Qwen1.5-0.5B-Chat-GGUF/resolve/main/qwen1_5-0_5b-chat-q3_k_m.gguf" selected>Qwen/Qwen1.5-0.5B-Chat Q3_K_M (350 MB)</option> -->
          <option value="https://huggingface.co/afrideva/TinyMistral-248M-SFT-v4-GGUF/resolve/main/tinymistral-248m-sft-v4.q8_0.gguf">tinymistral-248m-sft-v4 q8_0 (265.26 MB)</option>
          <option value="https://huggingface.co/TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF/resolve/main/tinyllama-1.1b-chat-v1.0.Q4_K_M.gguf">TinyLlama/TinyLlama-1.1B-Chat-v1.0 Q4_K_M (669 MB)</option>
          <option value="https://huggingface.co/Qwen/Qwen1.5-1.8B-Chat-GGUF/resolve/main/qwen1_5-1_8b-chat-q3_k_m.gguf">Qwen/Qwen1.5-1.8B-Chat Q3_K_M (1.02 GB)</option>
          <option value="https://huggingface.co/stabilityai/stablelm-2-zephyr-1_6b/resolve/main/stablelm-2-zephyr-1_6b-Q4_1.gguf">stabilityai/stablelm-2-zephyr-1_6b Q4_1 (1.07 GB)</option>
          <option value="https://huggingface.co/TKDKid1000/phi-1_5-GGUF/resolve/main/phi-1_5-Q4_K_M.gguf">microsoft/phi-1_5 Q4_K_M (918 MB)</option>
          <option value="https://huggingface.co/TheBloke/phi-2-GGUF/resolve/main/phi-2.Q3_K_M.gguf">microsoft/phi-2 Q3_K_M (1.48 GB)</option>
        </select>

        <label> Prompt: </label>

        <textarea id="prompt" name="prompt" rows="5">Suppose Alice originally had 3 apples, then Bob gave Alice 7 apples, then Alice gave Cook 5 apples, and then Tim gave Alice 3x the amount of apples Alice had. How many apples does Alice have now? Let’s think step by step.</textarea>

        <label> Result: </label>

        <!-- <textarea id="result" name="result" rows="10" autocomplete="off"></textarea> -->
        <pre id="result" name="result"></pre>
      </section>

      <section>
        <button id="run"> Run </button>
      </section>

      <section>
        <button id="run-progress-loading-model" aria-busy="true"hidden="hidden"> Loading model... </button>
        <button id="run-progress-loaded-model" aria-busy="true" hidden="hidden"> Loaded model </button>
        <button id="run-progress-generating" aria-busy="true" hidden="hidden"> Generating... </button>
      </section>
      
      <section>
        <progress id="model-progress" hidden="hidden" />
      </section>
    </main>
    
    <!-- example -->
    <script type="module" src="example-multi-thread.js?v=240213-5"></script>
  </body>
</html>