From ae6e0cfc41252e394e775427c87fb291e4e78b28 Mon Sep 17 00:00:00 2001
From: yumoqing <yumoqing@gmail.com>
Date: Wed, 4 Sep 2024 18:53:44 +0800
Subject: [PATCH] bugfix

---
 rtcllm/rtc.py | 4 ++--
 rtcllm/stt.py | 9 +++++----
 2 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/rtcllm/rtc.py b/rtcllm/rtc.py
index fe28b5c..a79724a 100644
--- a/rtcllm/rtc.py
+++ b/rtcllm/rtc.py
@@ -44,7 +44,7 @@ RTCPeerConnection.get_local_candidates = pc_get_local_candidates
 
 class RTCLLM:
 	def __init__(self, ws_url, iceServers):
-		self.stt_model = WhisperModel('large-v3', device="cuda", compute_type="float16")
+		# self.stt_model = WhisperModel('large-v3', device="cuda", compute_type="float16")
 		self.ws_url = ws_url
 		self.iceServers = iceServers
 		self.peers = DictObject()
@@ -101,7 +101,7 @@ class RTCLLM:
 		self.onlineList = data.onlineList
 	
 	async def vad_voiceend(self, peer, audio):
-		ret = await asr(self.stt_model, audio)
+		ret = await asr(audio)
 		# await peer.dc.send(txt)
 	
 
diff --git a/rtcllm/stt.py b/rtcllm/stt.py
index 57a7187..0ffee7b 100644
--- a/rtcllm/stt.py
+++ b/rtcllm/stt.py
@@ -30,11 +30,11 @@ opts = {
 	"asr":desc
 }
 
-async def asr(model, a_file):
-	"""
+async def asr(a_file):
 	oc  = OAuthClient(DictObject(**opts))
-	r = await oc("http://open-computing.cn", "asr", {"b64audio":audio})
-	print(f'{r=}')
+	with open(a_file, 'rb') as f:
+		r = await oc("http://open-computing.cn", "asr", {"audio_file":f})
+		print(f'{r=}')
 	"""
 	segments, info = model.transcribe(a_file, beam_size=5)
 	txt = ''
@@ -44,3 +44,4 @@ async def asr(model, a_file):
 		'content': txt,
 		'language': info.language
 	}
+	"""