feat(community): Add OpenAI Whisper options (#7650)

Co-authored-by: jacoblee93 <jacoblee93@gmail.com>
langchain-ai · Feb 6, 2025 · 0eedecf · 0eedecf
1 parent ded94d5
commit 0eedecf
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 1 deletion.
diff --git a/examples/src/document_loaders/openai_whisper_audio.ts b/examples/src/document_loaders/openai_whisper_audio.ts
@@ -2,7 +2,11 @@ import { OpenAIWhisperAudio } from "@langchain/community/document_loaders/fs/ope
 
 const filePath = "./src/document_loaders/example_data/test.mp3";
 
-const loader = new OpenAIWhisperAudio(filePath);
+const loader = new OpenAIWhisperAudio(filePath, {
+  transcriptionCreateParams: {
+    language: "en",
+  },
+});
 
 const docs = await loader.load();
 

diff --git a/libs/langchain-community/src/document_loaders/fs/openai_whisper_audio.ts b/libs/langchain-community/src/document_loaders/fs/openai_whisper_audio.ts
@@ -18,14 +18,18 @@ const MODEL_NAME = "whisper-1";
 export class OpenAIWhisperAudio extends BufferLoader {
   private readonly openAIClient: OpenAIClient;
 
+  private readonly transcriptionCreateParams?: Partial<OpenAIClient.Audio.TranscriptionCreateParams>;
+
   constructor(
     filePathOrBlob: string | Blob,
     fields?: {
       clientOptions?: ClientOptions;
+      transcriptionCreateParams?: Partial<OpenAIClient.Audio.TranscriptionCreateParams>;
     }
   ) {
     super(filePathOrBlob);
     this.openAIClient = new OpenAIClient(fields?.clientOptions);
+    this.transcriptionCreateParams = fields?.transcriptionCreateParams ?? {};
   }
 
   protected async parse(
@@ -38,6 +42,7 @@ export class OpenAIWhisperAudio extends BufferLoader {
       await this.openAIClient.audio.transcriptions.create({
         file: await toFile(raw, fileName),
         model: MODEL_NAME,
+        ...this.transcriptionCreateParams,
       });
     const document = new Document({
       pageContent: transcriptionResponse.text,