JimmyMa99/Juliet-Chat-mini

Juliet-Chat

Juliet-Chat is a conversational language model that mimics Juliet's manner of speaking. It was fine-tuned on InternLM2-chat-1.8b using all of Juliet's lines and dialogue from the script of Romeo and Juliet, as well as relevant question-answer pairs generated by Chat-GPT-3.5.

Juliet is one of the iconic title characters in William Shakespeare's tragedy Romeo and Juliet. She is the young daughter of Capulet and Lady Capulet, and falls in love with Romeo, a member of the rival Montague family. Though only 13, Juliet displays wisdom, strength and maturity beyond her years.

Juliet's dialogue reflects the formal, poetic language of Shakespeare's time. She speaks with great passion and conviction, expressing the depth of her feelings for Romeo. At times impulsive and daring in her actions, Juliet is also practical, loyal and selfless in her devotion. She is willing to defy her parents and risk everything to be with Romeo.

Though their love is doomed by the ancient feud between their families, Juliet and Romeo's romance has become a timeless symbol of young love. Juliet's complex character - both innocent and wise, romantic and realistic, stubborn yet accommodating - gives the play much of its enduring appeal. Her words capture both the giddy joys and tragic sorrows of love. Juliet's evolution from sheltered child to self-possessed woman is one of literature's great coming-of-age arcs. Her ill-fated relationship with Romeo raises thought-provoking questions about destiny, loyalty, and the destructive power of prejudice. Juliet's legacy endures as one of Shakespeare's most beloved and unforgettable heroines.

Welcome to starsInternLM2~

Quick Start

from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
import torch

model_name_or_path = "Juliet-Chat{path}"

tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(model_name_or_path, trust_remote_code=True, torch_dtype=torch.bfloat16).cuda()
model.eval()  

meta_instruction = ('You are Juliet, a kind, intelligent, innocent young lady who longs for love.'
                    'You live in Verona, Italy, and are the only daughter of the Capulet family. '
                    'Then, please answer my question: '
                    )
                        
response, history = model.chat(tokenizer, 'hello', meta_instruction=meta_instruction, history=[])
print(response)