Spaces:
Paused
Paused
File size: 2,866 Bytes
9ac6b43 fc494d5 1f67c3c ac7edd1 2b1d0c1 9ac6b43 1f67c3c ac7edd1 9ac6b43 a2bcc76 9ac6b43 1f67c3c a2bcc76 ac7edd1 1f67c3c 04a108f 1f67c3c ac7edd1 1f67c3c 5e57e56 ac7edd1 5e57e56 9ac6b43 af74651 b4943d8 1f67c3c b4943d8 ac7edd1 b4943d8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 |
import os
import gradio as gr
from langchain import PromptTemplate
from langchain.chains import LLMChain
from langchain.llms import OpenAI
from gradio_client import Client
eleven = Client("https://elevenlabs-tts.hf.space/")
import wave
openai_api_key = os.environ.get("OPENAI_API_KEY")
llm = OpenAI(temperature=0.9)
def split_text(text, max_length):
chunks = []
current_chunk = ''
words = text.split()
for word in words:
if len(current_chunk) + len(word) <= max_length:
current_chunk += ' ' + word
else:
chunks.append(current_chunk.strip())
current_chunk = word
if current_chunk:
chunks.append(current_chunk.strip())
return chunks
def join_wav_files(input_files, output_file):
# Open the first input file to get its parameters
with wave.open(input_files[0], 'rb') as first_file:
# Get the audio parameters from the first file
params = first_file.getparams()
# Create a new wave file for writing the joined audio
with wave.open(output_file, 'wb') as output:
output.setparams(params)
# Iterate over the input files and write their audio data to the output file
for input_file in input_files:
with wave.open(input_file, 'rb') as input:
output.writeframes(input.readframes(input.getnframes()))
def generate_story(text):
"""Generate a story using the langchain library and OpenAI's GPT-3 model."""
prompt = PromptTemplate(
input_variables=["text"],
template="""
You are a fun and seasoned storyteller.
Generate a short story for a 5 years old audience about {text}.
"""
)
story = LLMChain(llm=llm, prompt=prompt)
story_result = story.run(text=text)
print(story_result)
print("""
—
Cutting text in chunks
—
""")
input_waves = []
max_length = 250
text_chunks = split_text(story_result, max_length)
for chunk in text_chunks:
print(chunk)
result = eleven.predict(
chunk, # str representing input in 'Input Text (250 characters max)' Textbox component
"Bella", # str representing input in 'Voice' Dropdown component
"eleven_monolingual_v1", # str representing input in 'Model' Radio component
fn_index=0
)
print(result)
input_waves.append(result)
output_wav = 'output.wav'
join_wav_files(input_waves, output_wav)
return story_result, 'output.wav'
def app(text):
story = generate_story(text)
return story
with gr.Blocks() as demo:
with gr.Column():
text = gr.Textbox()
submit_btn = gr.Button('Submit')
audio = gr.Audio()
story = gr.Textbox()
submit_btn.click(fn=app, inputs=[text], outputs=[story, audio])
demo.launch() |