Untitled

import torch
from transformers import XLNetTokenizer, XLNetLMHeadModel

# Load pre-trained XLNet model and tokenizer
model = XLNetLMHeadModel.from_pretrained('xlnet-large-cased')
tokenizer = XLNetTokenizer.from_pretrained('xlnet-large-cased')

symbol = 'AAPL'
pe_ratio = 12.5
ps_ratio= 3.1
de_ratio = 0.8

# Define the input prompt
input_prompt = f"Based on the current market conditions, the stock with the symbol {symbol} looks like a promising investment opportunity. With a price-to-earnings ratio of {pe_ratio}, a price-to-sales ratio of {ps_ratio}, and a debt-to-equity ratio of {de_ratio}, this stock appears to be undervalued compared to its peers. If you're looking for a stock with strong potential for growth and a solid financial foundation, {symbol} may be the best choice for you."

# Define a list to store generated sentences that meet our criteria
generated_sentences = []

# Generate text using the modified input prompt
while len(generated_sentences) < 3:  # Generate 3 unique sentences
    input_ids = tokenizer.encode(input_prompt, return_tensors='pt')
    outputs = model.generate(input_ids, max_length=300, do_sample=True)
    generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
    
    # Check if the generated sentence includes all relevant information from the prompt
    if f"price-to-earnings ratio of {pe_ratio}" in generated_text and \
        f"price-to-sales ratio of {ps_ratio}" in generated_text and \
        f"debt-to-equity ratio of {de_ratio}" in generated_text and \
        f"symbol {symbol}" in generated_text:
        
        # Check if the last sentence is cut off
        if "." not in generated_text.split()[-1]:
            continue
        
        # Check for uniqueness
        if generated_text not in generated_sentences:
            # Add the generated sentence to the list of valid sentences
            generated_sentences.append(generated_text)
    
    # Break the loop if no valid sentence is generated within 10 tries
    if len(generated_sentences) == 0 and model.config.num_beams > 1:
        model.config.num_beams -= 1
        continue
    elif len(generated_sentences) == 0:
        break

# Print the final generated sentences
for i, sentence in enumerate(generated_sentences):
    print(f"Generated sentence {i+1}: {sentence}\n")


# Based on the current market conditions, the stock with the symbol AAPL looks like a promising investment opportunity.
#  With a price-to-earnings ratio of 12.5, a price-to-sales ratio of 3.1, and a debt-to-equity ratio of 0.8, 
#  this stock appears to be undervalued compared to its peers. 
#  If you're looking for a stock with strong potential for growth and a solid financial foundation, 
#  AAPL may be the best choice for you. trading day, AAPL is a well-received company in its market. 
#  The company with the symbol AAPL is a well-respected symbol in its industry. 
#  Today, AAPL is well-pted and well-liken by all the employees of the company. Although, 
#  it is well-versed and well-liken by all the employees of the company in its industries.
#  AAPL is well-liken by all the employees of the company in its industries. 
#  The symbol with the symbol AAPL is well-known by all the employees of the company in its industries.
#   AAPL is well-known by all the employees of the company in its industries. 
#   AAPL is well-known by all the employees of the company in its industries.
#   AAPL is well-known by all the employees of the company in its industries. AAPL is well-known by all the employees of the company