SiberianPersonaFred-2 Open-source Dialogue Generation Model - Effortlessly Simulate Conversations with Specific Personalities

Siberianpersonafred 2

Developed by SiberiaSoft

SiberianPersonaFred-2 is a dialogue generation model based on FRED-T5-XL, designed to simulate conversations with specific personalities.

Large Language Model

Transformers

OtherOpen Source License:MIT #Russian Dialogue Generation #Personalized Character Simulation #FRED-T5 Architecture

Downloads 175

Release Time : 9/17/2023

Model Overview

This model focuses on generating dialogue responses with personalized characteristics, supports Russian, and can be used to build personalized chatbots.

Model Features

Personalized Dialogue Generation

The model can generate dialogue responses that match specific personality traits based on preset descriptions.

Multi-turn Dialogue Support

Supports processing multi-turn dialogue contexts to maintain conversation coherence.

Russian Language Optimization

Specifically optimized for Russian dialogue.

Model Capabilities

Personalized Text Generation

Multi-turn Dialogue Processing

Russian Dialogue Understanding and Generation

Use Cases

Chatbots

Personalized Customer Service Assistant

Building a customer service chatbot with specific personality traits

Provides a more natural and personalized customer service experience

Social Companion Bot

Creating a virtual social companion with distinct personality characteristics

Offers more engaging and personalized conversation experiences

Education

Language Learning Assistant

Serving as a conversation partner for Russian language learners

Provides a natural Russian dialogue environment

🚀 SiberiaSoft/SiberianPersonaFred

This model is designed to imitate personalities in dialogues. For more details, click here.

This model is based on FRED-T5-XL

🚀 Quick Start

Personality Description Format

I'm a very intelligent girl and want to help my friends with useful advice.
I'm a guy, a consultant on various issues. I'm very smart and love to help my interlocutors.

You can also insert personal facts such as full name, age, etc. into the prompt:

I'm an 18 - year - old girl. I'm studying at the institute. I live with my parents. I have a cat. I'm looking for a boyfriend to start a family.

Article on habr: Link

💻 Usage Examples

Basic Usage

import torch
import transformers
use_cuda = torch.cuda.is_available()
device = torch.device("cuda" if use_cuda else "cpu")
t5_tokenizer = transformers.GPT2Tokenizer.from_pretrained("SiberiaSoft/SiberianPersonaFred-2")
t5_model = transformers.T5ForConditionalGeneration.from_pretrained("SiberiaSoft/SiberianPersonaFred-2")
while True:
    print('-'*80)
    dialog = []
    while True:
        msg = input('H:> ').strip()
        if len(msg) == 0:
            break
        msg = msg[0].upper() + msg[1:]
        dialog.append('Ты: ' + msg)
        # At the beginning, insert the character prompt.
        prompt = '<SC6>Я парень, консультант по разным вопросам. Я очень умный. Я люблю помогать собеседнику. Недавно, у меня был следующий диалог:' + '\n'.join(dialog) + '\nЯ: <extra_id_0>'
        input_ids = t5_tokenizer(prompt, return_tensors='pt').input_ids
        out_ids = t5_model.generate(input_ids=input_ids.to(device), do_sample=True, temperature=0.9, max_new_tokens=512, top_p=0.85,
                                      top_k=2, repetition_penalty=1.2)
        t5_output = t5_tokenizer.decode(out_ids[0][1:])
        if '</s>' in t5_output:
            t5_output = t5_output[:t5_output.find('</s>')].strip()
        t5_output = t5_output.replace('<extra_id_0>', '').strip()
        t5_output = t5_output.split('Собеседник')[0].strip()
        print('B:> {}'.format(t5_output))
        dialog.append('Я: ' + t5_output)

📄 License

This project is licensed under the MIT license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご