Text Generation
Transformers
English
Russian
legal
CyberFuture-3 / README.md
SkillForge45's picture
Update README.md
e495836 verified
---
license: mit
datasets:
- ParlAI/blended_skill_talk
- convai-challenge/conv_ai_2
- allenai/social_i_qa
language:
- en
- ru
metrics:
- accuracy
- precision
- recall
- perplexity
- bleu
- rouge
base_model:
- SkillForge45/CyberFuture-1
new_version: SkillForge45/CyberFuture-2
pipeline_tag: text-generation
library_name: transformers
tags:
- legal
---
# CyberFuture - Chat With AI
Uptades:
**-Web Search**
# Installation
1. Clone the model:
```bash
git clone https://huggingface.co/SkillForge45/CyberFuture-3
```
2. Install all requirements:
```bash
pip install torch transformers datasets googlesearch-python pyttsx3 speechrecognition fastapi uvicorn
```
# Usage
**METHOD #1 (HTML web interface):**
1. Launch the server:
```bash
python app.py
```
2. Start the HTML file (html_main.html)
**METHOD #2 (Python Uasge)**
Use this Python code:
```python
from model import ChatBot
# Initialize the chatbot
bot = ChatBot()
# Optionally train the model (requires GPU for good performance)
bot.train(epochs=3)
# Chat with web search
response = bot.generate_response("What's the latest news about AI?", use_web=True)
print(response)
# Voice interaction (requires microphone)
bot.voice_interface.speak(response)
user_input = bot.voice_interface.listen()
```
**METHOD #3 (Console)**
1. Text Chat:
```bash
curl -X POST "http://localhost:8000/chat/" \
-H "Content-Type: application/x-www-form-urlencoded" \
-d "prompt=What's the weather in London today?&use_web=true" #web search, if you not need web: use_web=False
```
2. Voice Chat:
```bash
curl -X POST "http://localhost:8000/chat/" \
-F "audio_file=@your_recording.wav" \
-F "use_web=true" \
-F "use_voice=true"
```
**METHOD 4 (Server)**
Start the app.py and go to http://localhost:8000/chat
# License
This model is licensed, see the LICENSE for more information