Spaces:
Runtime error
Runtime error
Commit ·
94a764b
1
Parent(s): a2538e6
change tts to Azure
Browse files- app.py +23 -5
- requirements.txt +1 -1
app.py
CHANGED
|
@@ -2,17 +2,34 @@
|
|
| 2 |
import gradio as gr
|
| 3 |
import openai, subprocess
|
| 4 |
from pathlib import Path
|
| 5 |
-
import
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
|
| 7 |
# 换成你自己的api_key
|
| 8 |
-
openai.api_key = "sk-
|
| 9 |
debatestyle = " "
|
| 10 |
messages = [{"role": "system", "content": '你是一名逻辑性很强的资深辩手,你擅长通过数据论据和学理论据来反驳,你的反驳总是一阵见血,而且很有逻辑性。接下来我会提出我的观点,你需要做的就是针锋相对地反驳我'}]
|
| 11 |
chat_transcript = ""
|
| 12 |
def transcribe(audio):
|
| 13 |
global messages
|
| 14 |
global chat_transcript
|
| 15 |
-
|
| 16 |
myfile=Path(audio)
|
| 17 |
myfile=myfile.rename(myfile.with_suffix('.wav'))
|
| 18 |
audio_file = open(myfile,"rb")
|
|
@@ -26,8 +43,9 @@ def transcribe(audio):
|
|
| 26 |
# print(response)
|
| 27 |
messages.append(system_message)
|
| 28 |
|
| 29 |
-
|
| 30 |
|
|
|
|
| 31 |
|
| 32 |
chat_transcript = ""
|
| 33 |
for message in messages:
|
|
@@ -52,7 +70,7 @@ def initway(api_key,style):
|
|
| 52 |
with gr.Blocks(css="#chatbot{height:300px} .overflow-y-auto{height:500px}") as init:
|
| 53 |
with gr.Row():
|
| 54 |
api_key = gr.Textbox(
|
| 55 |
-
lines=1, placeholder="api_key Here...", label="api_key",value="")
|
| 56 |
style = gr.Textbox(
|
| 57 |
lines=1, placeholder="style Here...", label="辩风" ,value="你是一名逻辑性很强的资深辩手,你擅长通过数据论据和学理论据来反驳,你的反驳总是一阵见血,而且很有逻辑性。接下来我会提出我的观点,你需要做的就是针锋相对地反驳我,每次回答不超过100个字")
|
| 58 |
btn = gr.Button(value="初始化")
|
|
|
|
| 2 |
import gradio as gr
|
| 3 |
import openai, subprocess
|
| 4 |
from pathlib import Path
|
| 5 |
+
import os
|
| 6 |
+
import azure.cognitiveservices.speech as speechsdk
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
# Azure tts
|
| 10 |
+
# This example requires environment variables named "SPEECH_KEY" and "SPEECH_REGION"
|
| 11 |
+
SPEECH_REGION='eastus'
|
| 12 |
+
SPEECH_KEY= 'af2d4aa2348b4b73b60487c73e0eb431'
|
| 13 |
+
|
| 14 |
+
speech_config = speechsdk.SpeechConfig(subscription='af2d4aa2348b4b73b60487c73e0eb431', region='eastus')
|
| 15 |
+
audio_config = speechsdk.audio.AudioOutputConfig(use_default_speaker=True)
|
| 16 |
+
|
| 17 |
+
# The language of the voice that speaks.
|
| 18 |
+
speech_config.speech_synthesis_voice_name=' zh-CN-XiaoxiaoNeural'
|
| 19 |
+
|
| 20 |
+
speech_synthesizer = speechsdk.SpeechSynthesizer(speech_config=speech_config, audio_config=audio_config)
|
| 21 |
+
|
| 22 |
+
|
| 23 |
|
| 24 |
# 换成你自己的api_key
|
| 25 |
+
openai.api_key = "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"
|
| 26 |
debatestyle = " "
|
| 27 |
messages = [{"role": "system", "content": '你是一名逻辑性很强的资深辩手,你擅长通过数据论据和学理论据来反驳,你的反驳总是一阵见血,而且很有逻辑性。接下来我会提出我的观点,你需要做的就是针锋相对地反驳我'}]
|
| 28 |
chat_transcript = ""
|
| 29 |
def transcribe(audio):
|
| 30 |
global messages
|
| 31 |
global chat_transcript
|
| 32 |
+
global speech_synthesizer
|
| 33 |
myfile=Path(audio)
|
| 34 |
myfile=myfile.rename(myfile.with_suffix('.wav'))
|
| 35 |
audio_file = open(myfile,"rb")
|
|
|
|
| 43 |
# print(response)
|
| 44 |
messages.append(system_message)
|
| 45 |
|
| 46 |
+
speech_synthesis_result = speech_synthesizer.speak_text_async(system_message['content']).get()
|
| 47 |
|
| 48 |
+
# 原windows tts 解决方案 subprocess.call(["wsay", system_message['content']])
|
| 49 |
|
| 50 |
chat_transcript = ""
|
| 51 |
for message in messages:
|
|
|
|
| 70 |
with gr.Blocks(css="#chatbot{height:300px} .overflow-y-auto{height:500px}") as init:
|
| 71 |
with gr.Row():
|
| 72 |
api_key = gr.Textbox(
|
| 73 |
+
lines=1, placeholder="api_key Here...", label="api_key",value="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")
|
| 74 |
style = gr.Textbox(
|
| 75 |
lines=1, placeholder="style Here...", label="辩风" ,value="你是一名逻辑性很强的资深辩手,你擅长通过数据论据和学理论据来反驳,你的反驳总是一阵见血,而且很有逻辑性。接下来我会提出我的观点,你需要做的就是针锋相对地反驳我,每次回答不超过100个字")
|
| 76 |
btn = gr.Button(value="初始化")
|
requirements.txt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
gradio==3.44.0
|
| 2 |
openai==0.28.0
|
| 3 |
-
|
|
|
|
| 1 |
gradio==3.44.0
|
| 2 |
openai==0.28.0
|
| 3 |
+
azure-cognitiveservices-speech==1.32.1
|