Fanny1366 commited on
Commit
90d9abf
·
verified ·
1 Parent(s): 7f4f62a

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +8 -5
app.py CHANGED
@@ -14,15 +14,15 @@ def img2text(url):
14
 
15
  # text2story
16
  def text2story(text):
17
- story_pipeline = pipeline("text-generation", model="facebook/opt-125m")
18
- # 调用 pipeline 生成故事文本
19
  result = story_pipeline(text, max_length=200, num_return_sequences=1)
20
  story_text = result[0]['generated_text']
21
  return story_text
22
 
23
  # text2audio
24
  def text2audio(story_text):
25
- # 使用 gTTS 库将文本转换为音频
26
  tts = gTTS(text=story_text, lang='en')
27
  # 创建一个内存中的字节流对象,用于存储音频数据
28
  audio_file = io.BytesIO()
@@ -30,7 +30,7 @@ def text2audio(story_text):
30
  tts.write_to_fp(audio_file)
31
  # 将文件指针移动到文件开头,以便后续读取
32
  audio_file.seek(0)
33
- return audio_file
34
 
35
  # main part
36
  st.set_page_config(page_title="Your Image to Audio Story",
@@ -63,4 +63,7 @@ if uploaded_file is not None:
63
 
64
  # Play button
65
  if st.button("Play Audio"):
66
- st.audio(audio_data, format="audio/mpeg")
 
 
 
 
14
 
15
  # text2story
16
  def text2story(text):
17
+ # 使用 Hugging Face 的 text-generation 模型生成故事
18
+ story_pipeline = pipeline("text-generation", model="agentica-org/DeepScaleR-1.5B-Preview")
19
  result = story_pipeline(text, max_length=200, num_return_sequences=1)
20
  story_text = result[0]['generated_text']
21
  return story_text
22
 
23
  # text2audio
24
  def text2audio(story_text):
25
+ # 使用 gTTS 将文本转换为音频
26
  tts = gTTS(text=story_text, lang='en')
27
  # 创建一个内存中的字节流对象,用于存储音频数据
28
  audio_file = io.BytesIO()
 
30
  tts.write_to_fp(audio_file)
31
  # 将文件指针移动到文件开头,以便后续读取
32
  audio_file.seek(0)
33
+ return {'audio': audio_file, 'sampling_rate': 16000} # 返回音频数据和采样率
34
 
35
  # main part
36
  st.set_page_config(page_title="Your Image to Audio Story",
 
63
 
64
  # Play button
65
  if st.button("Play Audio"):
66
+ st.audio(audio_data['audio'],
67
+ format="audio/wav",
68
+ start_time=0,
69
+ sample_rate=audio_data['sampling_rate'])