feat(providers): add transformers provider

package.json

@@ -13,6 +13,7 @@
   "dependencies": {
     "@headlessui/react": "^2.2.0",
     "@headlessui/tailwindcss": "^0.2.2",
+    "@huggingface/transformers": "^3.7.5",
     "@iarna/toml": "^2.2.5",
     "@icons-pack/react-simple-icons": "^12.3.0",
     "@langchain/anthropic": "^0.3.24",
@@ -24,7 +25,6 @@
     "@langchain/openai": "^0.6.2",
     "@langchain/textsplitters": "^0.1.0",
     "@tailwindcss/typography": "^0.5.12",
-    "@xenova/transformers": "^2.17.2",
     "axios": "^1.8.3",
     "better-sqlite3": "^11.9.1",
     "clsx": "^2.1.0",

src/lib/huggingfaceTransformer.ts

@@ -1,5 +1,6 @@
 import { Embeddings, type EmbeddingsParams } from '@langchain/core/embeddings';
 import { chunkArray } from '@langchain/core/utils/chunk_array';
+import { pipeline } from '@huggingface/transformers';
 
 export interface HuggingFaceTransformersEmbeddingsParams
   extends EmbeddingsParams {
@@ -67,12 +68,7 @@ export class HuggingFaceTransformersEmbeddings
   }
 
   private async runEmbedding(texts: string[]) {
-    const { pipeline } = await import('@xenova/transformers');
-
-    const pipe = await (this.pipelinePromise ??= pipeline(
-      'feature-extraction',
-      this.model,
-    ));
+    const pipe = await pipeline('feature-extraction', this.model);
 
     return this.caller.call(async () => {
       const output = await pipe(texts, { pooling: 'mean', normalize: true });

src/lib/models/providers/index.ts

@@ -2,10 +2,12 @@ import { ModelProviderUISection } from '@/lib/config/types';
 import { ProviderConstructor } from './baseProvider';
 import OpenAIProvider from './openai';
 import OllamaProvider from './ollama';
+import TransformersProvider from './transformers';
 
 export const providers: Record<string, ProviderConstructor<any>> = {
   openai: OpenAIProvider,
   ollama: OllamaProvider,
+  transformers: TransformersProvider,
 };
 
 export const getModelProvidersUIConfigSection =

src/lib/models/providers/transformers.ts

@@ -0,0 +1,88 @@
+import { BaseChatModel } from '@langchain/core/language_models/chat_models';
+import { Model, ModelList, ProviderMetadata } from '../types';
+import BaseModelProvider from './baseProvider';
+import { Embeddings } from '@langchain/core/embeddings';
+import { UIConfigField } from '@/lib/config/types';
+import { getConfiguredModelProviderById } from '@/lib/config/serverRegistry';
+import { HuggingFaceTransformersEmbeddings } from '@/lib/huggingfaceTransformer';
+
+interface TransformersConfig {}
+
+const defaultEmbeddingModels: Model[] = [
+  {
+    name: 'all-MiniLM-L6-v2',
+    key: 'Xenova/all-MiniLM-L6-v2',
+  },
+  {
+    name: 'mxbai-embed-large-v1',
+    key: 'mixedbread-ai/mxbai-embed-large-v1',
+  },
+  {
+    name: 'nomic-embed-text-v1',
+    key: 'Xenova/nomic-embed-text-v1',
+  },
+];
+
+const providerConfigFields: UIConfigField[] = [];
+
+class TransformersProvider extends BaseModelProvider<TransformersConfig> {
+  constructor(id: string, name: string, config: TransformersConfig) {
+    super(id, name, config);
+  }
+
+  async getDefaultModels(): Promise<ModelList> {
+    return {
+      embedding: [...defaultEmbeddingModels],
+      chat: [],
+    };
+  }
+
+  async getModelList(): Promise<ModelList> {
+    const defaultModels = await this.getDefaultModels();
+    const configProvider = getConfiguredModelProviderById(this.id)!;
+
+    return {
+      embedding: [
+        ...defaultModels.embedding,
+        ...configProvider.embeddingModels,
+      ],
+      chat: [],
+    };
+  }
+
+  async loadChatModel(key: string): Promise<BaseChatModel> {
+    throw new Error('Transformers Provider does not support chat models.');
+  }
+
+  async loadEmbeddingModel(key: string): Promise<Embeddings> {
+    const modelList = await this.getModelList();
+    const exists = modelList.embedding.find((m) => m.key === key);
+
+    if (!exists) {
+      throw new Error(
+        'Error Loading OpenAI Embedding Model. Invalid Model Selected.',
+      );
+    }
+
+    return new HuggingFaceTransformersEmbeddings({
+      model: key,
+    });
+  }
+
+  static parseAndValidate(raw: any): TransformersConfig {
+    return {};
+  }
+
+  static getProviderConfigFields(): UIConfigField[] {
+    return providerConfigFields;
+  }
+
+  static getProviderMetadata(): ProviderMetadata {
+    return {
+      key: 'transformers',
+      name: 'Transformers',
+    };
+  }
+}
+
+export default TransformersProvider;