andrei-stoica
/
ai_sandbox
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

cleaning up backend sins now

This commit is contained in:
Andrei Stoica 2024-02-26 11:52:13 -05:00
parent 64bb9f9db3
commit 1916185f19
1 changed files with 6 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
speech-speech/backend/api.py
View File

@ -1,8 +1,8 @@
from openai import OpenAI from openai import OpenAI
from fastapi import FastAPI, File, Response, Request from fastapi import FastAPI, File, Response, Request
from fastapi.middleware.cors import CORSMiddleware from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import FileResponse
from pydantic import BaseModel from pydantic import BaseModel
from io import BytesIO
app = FastAPI() app = FastAPI()
@ -26,22 +26,20 @@ class Conversation(BaseModel):
@app.post("/get-text") @app.post("/get-text")
def stt(audio: bytes = File()): async def stt(audio: bytes = File()):
with open("audio.webm", "wb+") as f: with BytesIO(audio) as f:
f.write(audio) f.name = "audio.mp3"
transcript = openAI_clinet.audio.transcriptions.create( transcript = openAI_clinet.audio.transcriptions.create(
model="whisper-1", model="whisper-1",
file=f, file=f,
response_format="text", response_format="text",
) )
data = {"len": len(audio), "user-transcript": transcript} data = {"len": len(audio), "user-transcript": transcript}
return data return data
@app.post("/conversation") @app.post("/conversation")
async def get_next_response(request: Request): async def get_next_response(request: Request):
# role = "test"
# res_msg = "temp test response"
messages = await request.json() messages = await request.json()
res = openAI_clinet.chat.completions.create( res = openAI_clinet.chat.completions.create(
model="gpt-3.5-turbo", model="gpt-3.5-turbo",
@ -57,10 +55,6 @@ async def get_next_response(request: Request):
@app.get("/speak") @app.get("/speak")
def tts(text: str): def tts(text: str):
res = openAI_clinet.audio.speech.create( res = openAI_clinet.audio.speech.create(
model="tts-1", model="tts-1", voice="nova", input=text, response_format="mp3"
voice="nova",
input=text,
response_format='mp3'
) )
# this works for now but I need to find a way to stream this to response
return Response(content=res.content, media_type="audio/mp3") return Response(content=res.content, media_type="audio/mp3")