refactor: repartition server-side and client-side code

2025-10-11 18:26:07 +08:00
parent 7acc4949fb
commit e9b573a276
309 changed files with 631 additions and 962 deletions
--- a/app/.server/modules/llm/providers/huggingface.ts
+++ b/app/.server/modules/llm/providers/huggingface.ts
@@ -0,0 +1,96 @@
+import { createOpenAI } from '@ai-sdk/openai';
+import type { LanguageModel } from 'ai';
+import { BaseProvider } from '~/.server/modules/llm/base-provider';
+import type { ModelInfo } from '~/.server/modules/llm/types';
+import type { IProviderSetting } from '~/types/model';
+
+export default class HuggingFaceProvider extends BaseProvider {
+  name = 'HuggingFace';
+  getApiKeyLink = 'https://huggingface.co/settings/tokens';
+
+  staticModels: ModelInfo[] = [
+    {
+      name: 'Qwen/Qwen2.5-Coder-32B-Instruct',
+      label: 'Qwen2.5-Coder-32B-Instruct (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: '01-ai/Yi-1.5-34B-Chat',
+      label: 'Yi-1.5-34B-Chat (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'codellama/CodeLlama-34b-Instruct-hf',
+      label: 'CodeLlama-34b-Instruct (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'NousResearch/Hermes-3-Llama-3.1-8B',
+      label: 'Hermes-3-Llama-3.1-8B (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'Qwen/Qwen2.5-Coder-32B-Instruct',
+      label: 'Qwen2.5-Coder-32B-Instruct (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'Qwen/Qwen2.5-72B-Instruct',
+      label: 'Qwen2.5-72B-Instruct (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'meta-llama/Llama-3.1-70B-Instruct',
+      label: 'Llama-3.1-70B-Instruct (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'meta-llama/Llama-3.1-405B',
+      label: 'Llama-3.1-405B (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: '01-ai/Yi-1.5-34B-Chat',
+      label: 'Yi-1.5-34B-Chat (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'codellama/CodeLlama-34b-Instruct-hf',
+      label: 'CodeLlama-34b-Instruct (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+    {
+      name: 'NousResearch/Hermes-3-Llama-3.1-8B',
+      label: 'Hermes-3-Llama-3.1-8B (HuggingFace)',
+      provider: 'HuggingFace',
+      maxTokenAllowed: 8000,
+    },
+  ];
+
+  getModelInstance(options: { model: string; providerSettings?: Record<string, IProviderSetting> }): LanguageModel {
+    const { model, providerSettings } = options;
+
+    const { apiKey } = this.getProviderBaseUrlAndKey(providerSettings?.[this.name]);
+
+    if (!apiKey) {
+      throw new Error(`Missing API key for ${this.name} provider`);
+    }
+
+    const openai = createOpenAI({
+      baseURL: 'https://api-inference.huggingface.co/v1/',
+      apiKey,
+    });
+
+    return openai(model);
+  }
+}