Spaces:
Sleeping
Sleeping
File size: 4,134 Bytes
672e79a |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 |
import gradio as gr
from transformers import pipeline
import requests
import os
import html
import base64
import random
import json
from datetime import datetime
chat_gen = pipeline("text-generation", model="tiiuae/falcon-rw-1b")
voice_options = {
"Rachel": "EXAVITQu4vr4xnSDxMaL",
"Adam": "21m00Tcm4TlvDq8ikWAM",
"Elli": "AZnzlk1XvdvUeBnXmlld",
"Josh": "VR6AewLTigWG4xSOukaG"
}
DEFAULT_ELEVEN_API_KEY = "sk_4e67c39c0e9cbc87462cd2643e1f4d1d9959d7d81203adc2"
library_file = "/tmp/library.json"
def generate_chat_bundle(prompt, *character_inputs):
characters = [c.split(":")[0].strip() for c in character_inputs if ":" in c]
voices = {c.split(":")[0].strip(): c.split(":")[1].strip() for c in character_inputs if ":" in c}
system_prompt = (
f"Group chat between {', '.join(characters)}. Prompt: {prompt}. "
"Use casual texting style: Name: message"
)
response = chat_gen(system_prompt, max_new_tokens=300)[0]["generated_text"]
chat_lines = [(line.split(':', 1)[0].strip(), line.split(':', 1)[1].strip()) for line in response.splitlines() if ':' in line]
html_chat = ""
file_paths = []
api_key = os.getenv("ELEVEN_API_KEY", DEFAULT_ELEVEN_API_KEY)
for i, (name, msg) in enumerate(chat_lines):
color = random.choice(["#DCF8C6", "#FEE2E2", "#EDEDED", "#C7E2FF"])
html_chat += f"""<div style='margin: 10px; text-align: {'left' if i % 2 == 0 else 'right'}'>
<div style='display:inline-block; padding: 10px 15px; background: {color}; border-radius: 15px; max-width: 70%;'>
<strong>{html.escape(name)}:</strong> {html.escape(msg)}
</div></div>"""
voice_id = voices.get(name, list(voice_options.values())[0])
eleven_response = requests.post(
f"https://api.elevenlabs.io/v1/text-to-speech/{voice_id}",
headers={
"xi-api-key": api_key,
"Content-Type": "application/json"
},
json={"text": msg, "model_id": "eleven_monolingual_v1"}
)
if eleven_response.status_code == 200:
file_path = f"/tmp/audio_{i}_{name}.mp3"
with open(file_path, 'wb') as f:
f.write(eleven_response.content)
file_paths.append(file_path)
timestamp = datetime.now().strftime('%Y%m%d-%H%M%S')
html_path = f"/tmp/fakechat_{timestamp}.html"
with open(html_path, "w") as f:
f.write(f"<html><body>{html_chat}</body></html>")
file_paths.append(html_path)
if os.path.exists(library_file):
with open(library_file, "r") as f:
lib = json.load(f)
else:
lib = []
lib.insert(0, {
"timestamp": timestamp,
"characters": characters,
"prompt": prompt,
"html": html_path,
"audio_files": file_paths[:-1]
})
with open(library_file, "w") as f:
json.dump(lib, f, indent=2)
return file_paths
def get_library():
if not os.path.exists(library_file):
return "No stories saved yet."
with open(library_file, "r") as f:
lib = json.load(f)
html_links = [f"<li><b>{item['timestamp']}</b>: {', '.join(item['characters'])} — <a href='file={item['html']}' target='_blank'>Chat</a></li>" for item in lib]
return "<ul>" + "".join(html_links) + "</ul>"
def build_character_inputs():
return [gr.Textbox(label=f"Character {i+1} (Format: Name:Voice)", placeholder="e.g. Anna:Rachel") for i in range(4)]
with gr.Blocks() as app:
with gr.Row():
with gr.Column(scale=3):
characters = build_character_inputs()
prompt = gr.Textbox(label="Scene Prompt")
generate_btn = gr.Button("🎤 Generate Chat + Audio")
file_output = gr.File(file_types=[".html", ".mp3"], label="Download Files")
with gr.Column(scale=1):
lib_html = gr.HTML(label="📁 My Saved Stories")
refresh_btn = gr.Button("🔁 Refresh Library")
generate_btn.click(fn=generate_chat_bundle, inputs=[prompt] + characters, outputs=file_output)
refresh_btn.click(fn=get_library, outputs=lib_html)
app.launch() |