mirror of
https://github.com/memohai/Memoh.git
synced 2026-04-27 07:16:19 +09:00
sheepbox8646
2 lines
6.0 KiB
JavaScript
2 lines
6.0 KiB
JavaScript
import{_ as t,o,c as r,ag as i}from"./chunks/framework.DXGyWiRo.js";const p=JSON.parse('{"title":"LLM Provider and Model","description":"","frontmatter":{},"headers":[],"relativePath":"getting-started/provider-and-model.md","filePath":"getting-started/provider-and-model.md","lastUpdated":1774787421000}'),a={name:"getting-started/provider-and-model.md"};function d(n,e,l,s,h,c){return o(),r("div",null,[...e[0]||(e[0]=[i('<h1 id="llm-provider-and-model" tabindex="-1">LLM Provider and Model <a class="header-anchor" href="#llm-provider-and-model" aria-label="Permalink to "LLM Provider and Model""></a></h1><p>To use Memoh, you first need to configure at least one LLM Provider and at least one Model.</p><h2 id="llm-provider" tabindex="-1">LLM Provider <a class="header-anchor" href="#llm-provider" aria-label="Permalink to "LLM Provider""></a></h2><p>An LLM Provider represents a connection to an AI service (like OpenAI, Anthropic, or a self-hosted compatible API). It stores the base URL and authentication credentials.</p><h3 id="creating-a-provider" tabindex="-1">Creating a Provider <a class="header-anchor" href="#creating-a-provider" aria-label="Permalink to "Creating a Provider""></a></h3><ol><li>Navigate to the <strong>Providers</strong> page from the sidebar.</li><li>Click the <strong>Add Provider</strong> button at the bottom of the sidebar.</li><li>Fill in the following fields: <ul><li><strong>Name</strong>: A display name for this provider (e.g., "OpenAI").</li><li><strong>Base URL</strong>: The root URL of the API (e.g., <code>https://api.openai.com/v1</code>).</li><li><strong>API Key</strong>: Your authentication token for the service.</li></ul></li><li>Click <strong>Create</strong>.</li></ol><h3 id="oauth-authentication" tabindex="-1">OAuth Authentication <a class="header-anchor" href="#oauth-authentication" aria-label="Permalink to "OAuth Authentication""></a></h3><p>Some providers (like OpenAI) support OAuth-based authentication. If a provider supports OAuth:</p><ol><li>Select the provider from the list.</li><li>Click <strong>Connect with OAuth</strong> in the provider settings form.</li><li>Follow the authorization flow in the popup window.</li><li>Once authorized, the provider will use the OAuth token instead of a manual API key.</li></ol><h3 id="import-models" tabindex="-1">Import Models <a class="header-anchor" href="#import-models" aria-label="Permalink to "Import Models""></a></h3><p>Memoh can automatically discover and import available models from a provider:</p><ol><li>Select a provider from the list.</li><li>Click <strong>Import Models</strong>.</li><li>Memoh will query the provider's API to fetch available models.</li><li>Select which models to import and click <strong>Import</strong>.</li></ol><p>This saves time compared to manually adding each model one by one.</p><h3 id="managing-providers" tabindex="-1">Managing Providers <a class="header-anchor" href="#managing-providers" aria-label="Permalink to "Managing Providers""></a></h3><ul><li><strong>Edit</strong>: Select a provider from the list and use the form on the right to update its name, URL, or API key.</li><li><strong>Test</strong>: Click <strong>Test Connection</strong> to verify the provider is reachable.</li><li><strong>Delete</strong>: Use the <strong>Delete Provider</strong> button in the provider settings form.</li></ul><hr><h2 id="model" tabindex="-1">Model <a class="header-anchor" href="#model" aria-label="Permalink to "Model""></a></h2><p>A Model is a specific AI instance (like <code>gpt-4o</code> or <code>text-embedding-3-small</code>) that belongs to a Provider. Memoh distinguishes between <strong>Chat</strong> models (for conversation) and <strong>Embedding</strong> models (for memory search).</p><h3 id="adding-a-model" tabindex="-1">Adding a Model <a class="header-anchor" href="#adding-a-model" aria-label="Permalink to "Adding a Model""></a></h3><ol><li>Select a Provider from the list on the <strong>Providers</strong> page.</li><li>Click <strong>Add Model</strong> in the model list section.</li><li>Configure the following fields:</li></ol><table tabindex="0"><thead><tr><th>Field</th><th>Required</th><th>Description</th></tr></thead><tbody><tr><td><strong>Type</strong></td><td>Yes</td><td><code>chat</code> for conversation, <code>embedding</code> for vector search.</td></tr><tr><td><strong>Model ID</strong></td><td>Yes</td><td>The exact identifier used by the provider (e.g., <code>gpt-4o</code>).</td></tr><tr><td><strong>Name</strong></td><td>No</td><td>A friendly display name (defaults to Model ID).</td></tr><tr><td><strong>Client Type</strong></td><td>Yes (Chat)</td><td>The API protocol: <code>openai-responses</code>, <code>openai-completions</code>, <code>anthropic-messages</code>, or <code>google-generative-ai</code>.</td></tr><tr><td><strong>Input Modalities</strong></td><td>Yes (Chat)</td><td>Capabilities supported: <code>text</code> (default), <code>image</code>, <code>audio</code>, <code>video</code>, <code>file</code>.</td></tr><tr><td><strong>Supports Reasoning</strong></td><td>No</td><td>Enable if the model supports internal reasoning steps (e.g., OpenAI o1).</td></tr><tr><td><strong>Dimensions</strong></td><td>Yes (Embed)</td><td>The vector size for embedding models (e.g., 1536).</td></tr></tbody></table><ol start="4"><li>Click <strong>Create</strong>.</li></ol><h3 id="managing-models" tabindex="-1">Managing Models <a class="header-anchor" href="#managing-models" aria-label="Permalink to "Managing Models""></a></h3><ul><li><strong>Edit</strong>: Click the edit icon next to a model in the list.</li><li><strong>Delete</strong>: Click the trash icon next to a model to remove it.</li></ul><hr><h2 id="next-steps" tabindex="-1">Next Steps <a class="header-anchor" href="#next-steps" aria-label="Permalink to "Next Steps""></a></h2><p>Now that you have configured your models, you can proceed to <a href="/getting-started/bot.html">Create and Configure a Bot</a>.</p>',27)])])}const m=t(a,[["render",d]]);export{p as __pageData,m as default};
|