Commit @4de8651f1af5c4ec909aac03230620b8da90c167 - yjyoon/whisper_streaming

Quentin Fuxa 2024-12-25

/ws to /asr to distinguish protocol ws:// from endpoint

@4de8651f1af5c4ec909aac03230620b8da90c167

6d3d6ca

4de8651

README.md

--- README.md

+++ README.md


 ### How the Live Interface Works
 
 - Once you **allow microphone access**, the page records small chunks of audio using the **MediaRecorder** API in **webm/opus** format.  
-- These chunks are sent over a **WebSocket** to the FastAPI endpoint at `/ws`.  
+- These chunks are sent over a **WebSocket** to the FastAPI endpoint at `/asr`.  
 - The Python server decodes `.webm` chunks on the fly using **FFmpeg** and streams them into the **whisper streaming** implementation for transcription.  
 - **Partial transcription** appears as soon as enough audio is processed. The “unvalidated” text is shown in **lighter or grey color** (i.e., an ‘aperçu’) to indicate it’s still buffered partial output. Once Whisper finalizes that segment, it’s displayed in normal text.  
 - You can watch the transcription update in near real time, ideal for demos, prototyping, or quick debugging.

6d3d6ca

4de8651

src/live_transcription.html

--- src/live_transcription.html

+++ src/live_transcription.html


       </div>
       <div>
         <label for="websocketInput">WebSocket URL:</label>
-        <input id="websocketInput" type="text" value="ws://localhost:8000/ws" />
+        <input id="websocketInput" type="text" value="ws://localhost:8000/asr" />
       </div>
     </div>
   </div>

       websocket,
       recorder,
       chunkDuration = 1000,
-      websocketUrl = "ws://localhost:8000/ws";
+      websocketUrl = "ws://localhost:8000/asr";
 
     // Tracks whether the user voluntarily closed the WebSocket
     let userClosing = false;

6d3d6ca

4de8651

whisper_fastapi_online_server.py

--- whisper_fastapi_online_server.py

+++ whisper_fastapi_online_server.py


     )
     return process
 
-@app.websocket("/ws")
+@app.websocket("/asr")
 async def websocket_endpoint(websocket: WebSocket):
     await websocket.accept()
     print("WebSocket connection opened.")

Add a comment

Open 0
Closed 0

List

...	...	@@ -86,7 +86,7 @@
86	86	### How the Live Interface Works
87	87
88	88	- Once you allow microphone access, the page records small chunks of audio using the MediaRecorder API in webm/opus format.
89		-- These chunks are sent over a WebSocket to the FastAPI endpoint at `/ws`.
	89	+- These chunks are sent over a WebSocket to the FastAPI endpoint at `/asr`.
90	90	- The Python server decodes `.webm` chunks on the fly using FFmpeg and streams them into the whisper streaming implementation for transcription.
91	91	- Partial transcription appears as soon as enough audio is processed. The “unvalidated” text is shown in lighter or grey color (i.e., an ‘aperçu’) to indicate it’s still buffered partial output. Once Whisper finalizes that segment, it’s displayed in normal text.
92	92	- You can watch the transcription update in near real time, ideal for demos, prototyping, or quick debugging.

...	...	@@ -92,7 +92,7 @@
92	92	</div>
93	93	<div>
94	94	<label for="websocketInput">WebSocket URL:</label>
95		- <input id="websocketInput" type="text" value="ws://localhost:8000/ws" />
	95	+ <input id="websocketInput" type="text" value="ws://localhost:8000/asr" />
96	96	</div>
97	97	</div>
98	98	</div>
...	...	@@ -105,7 +105,7 @@
105	105	websocket,
106	106	recorder,
107	107	chunkDuration = 1000,
108		- websocketUrl = "ws://localhost:8000/ws";
	108	+ websocketUrl = "ws://localhost:8000/asr";
109	109
110	110	// Tracks whether the user voluntarily closed the WebSocket
111	111	let userClosing = false;

...	...	@@ -57,7 +57,7 @@
57	57	)
58	58	return process
59	59
60		-@app.websocket("/ws")
	60	+@app.websocket("/asr")
61	61	async def websocket_endpoint(websocket: WebSocket):
62	62	await websocket.accept()
63	63	print("WebSocket connection opened.")

Delete comment