AssistGPT / model.py
shing12345's picture
Update model.py
d021341 verified
import torch
from transformers import BartForConditionalGeneration, BartTokenizer, GPT2LMHeadModel, GPT2Tokenizer
import argparse
import sys
class AdvancedSummarizer:
def __init__(self, model_name="facebook/bart-large-cnn"):
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
self.model = BartForConditionalGeneration.from_pretrained(model_name).to(self.device)
self.tokenizer = BartTokenizer.from_pretrained(model_name)
def summarize(self, text, max_length=150, min_length=50, length_penalty=2.0, num_beams=4):
inputs = self.tokenizer([text], max_length=1024, return_tensors="pt", truncation=True)
inputs = inputs.to(self.device)
summary_ids = self.model.generate(
inputs["input_ids"],
num_beams=num_beams,
max_length=max_length,
min_length=min_length,
length_penalty=length_penalty
)
summary = self.tokenizer.decode(summary_ids[0], skip_special_tokens=True)
return summary
def main_summarizer():
# Example usage
summarizer = AdvancedSummarizer()
text = """Your text here""" # Replace with your actual text
summary = summarizer.summarize(text)
print("Summary:")
print(summary)
class AdvancedTextGenerator:
def __init__(self, model_name="gpt2-medium"):
try:
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print(f"Using device: {self.device}")
self.model = GPT2LMHeadModel.from_pretrained(model_name).to(self.device)
self.tokenizer = GPT2Tokenizer.from_pretrained(model_name)
except Exception as e:
print(f"Error initializing the model: {e}")
sys.exit(1)
def generate_text(self, prompt, max_length=100, num_return_sequences=1,
temperature=1.0, top_k=50, top_p=0.95, repetition_penalty=1.0):
try:
input_ids = self.tokenizer.encode(prompt, return_tensors="pt").to(self.device)
output_sequences = self.model.generate(
input_ids=input_ids,
max_length=max_length + len(input_ids[0]),
temperature=temperature,
top_k=top_k,
top_p=top_p,
repetition_penalty=repetition_penalty,
do_sample=True,
num_return_sequences=num_return_sequences,
)
generated_sequences = []
for generated_sequence in output_sequences:
text = self.tokenizer.decode(generated_sequence, clean_up_tokenization_spaces=True)
total_sequence = text[len(self.tokenizer.decode(input_ids[0], clean_up_tokenization_spaces=True)):]
generated_sequences.append(total_sequence)
return generated_sequences
except Exception as e:
return [f"Error during text generation: {e}"]
def main_generator():
parser = argparse.ArgumentParser(description="Advanced Text Generator")
parser.add_argument("--prompt", type=str, help="Starting prompt for text generation")
parser.add_argument("--max_length", type=int, default=100, help="Maximum length of generated text")
parser.add_argument("--num_sequences", type=int, default=1, help="Number of sequences to generate")
parser.add_argument("--temperature", type=float, default=1.0, help="Temperature for sampling")
parser.add_argument("--top_k", type=int, default=50, help="Top-k sampling parameter")
parser.add_argument("--top_p", type=float, default=0.95, help="Top-p sampling parameter")
parser.add_argument("--repetition_penalty", type=float, default=1.0, help="Repetition penalty")
args = parser.parse_args()
generator = AdvancedTextGenerator()
if args.prompt:
prompt = args.prompt
else:
print("Please enter the prompt for text generation:")
prompt = input().strip()
generated_texts = generator.generate_text(
prompt,
max_length=args.max_length,
num_return_sequences=args.num_sequences,
temperature=args.temperature,
top_k=args.top_k,
top_p=args.top_p,
repetition_penalty=args.repetition_penalty
)
print("\nGenerated Text(s):")
for i, text in enumerate(generated_texts, 1):
print(f"\n--- Sequence {i} ---")
print(text)
if __name__ == "__main__":
main_summarizer() # Call the summarizer main function
main_generator() # Call the text generator main function