Setting up Simple Application to experiment with prompts

Step 1: Install Required Libraries: First, you'll need to install the Hugging Face transformers library and torch for PyTorch support. 

pip install transformers torch

Step 2: Load a Pre-trained Model: You can use Hugging Face's bloom-1b7 model, a popular model to generate human-like text across a wide range of languages. You can use it for creative writing, code generation, and more..

 

Here's a basic script to experiment with it.

 

promptSetup.py

import os
from transformers import BloomForCausalLM, BloomTokenizerFast

# Load the BLOOM model and tokenizer
model_name = "bigscience/bloom-1b7"
model = BloomForCausalLM.from_pretrained(model_name)
tokenizer = BloomTokenizerFast.from_pretrained(model_name)

def generate_response(prompt):
    # Tokenize the prompt
    inputs = tokenizer(prompt, return_tensors="pt")

    # Ensure input_ids are of type LongTensor (for embedding layer)
    input_ids = inputs["input_ids"].long()  # Explicitly cast to LongTensor

    # Generate a response with sampling enabled
    outputs = model.generate(input_ids, 
                             max_length=100, 
                             num_return_sequences=1, 
                             temperature=0.7,   # Temperature for sampling
                             do_sample=True)    # Enable sampling mode
    
    # Decode and return the result
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

# Function to format the conversation output
def format_conversation(response):
    separator = "\n" + "="*50 + "\n"  # You can adjust the separator length
    formatted_conversation = f"BLOOM: {response}\n"
    return formatted_conversation

# Interactive loop
print("Welcome to BLOOM-powered conversation!\n" + "="*50 + "\n")

while True:
    user_input = input("You: ")
    if user_input.lower() == "exit":
        print("\nEnding the conversation. Goodbye!")
        break
    response = generate_response(user_input)
    print(format_conversation(response))
    print("\n" + "-"*50 + "\n")  # Adding spacing between conversations

Execute following command to experiment with the Application.

python3 promptSetup.py

 

Prompt 1: For the input “Once Upon A Time”

You: Once Upon A Time

==================================================
BLOOM: Once Upon A Time In Hollywood (1968) - The first time I've seen a movie in my life, which I will never forget. The movie is a good one, but I can't even remember what it was about. I just remember the music. I just remembered the title and the poster. The film is called The Girl from the Sea, which I thought was a good name for the movie. I got in the car, and I was going to go to this movie. I called the theater

 

Prompt 2: For the input “SQL Stands for Structured”

You: SQL Stands for Structured

==================================================
BLOOM: SQL Stands for Structured Query Language. It is a set of SQL statements. SQL is used to access database. SQL is part of Oracle Database. SQL is a language for creating a query. SQL is a combination of the words Structured Query Language and Language. SQL is used to create queries. SQL is used to connect the database and database server. It is a language used to do database operations. The SQL statement is very powerful. It can create a query and connect to the database.

Prompt 3: For the Input “Tell Me a Joke in English”

You: Tell Me a Joke in English

==================================================
BLOOM: Tell Me a Joke in English (I don't know how to write that in English, so I don't know if that is a joke)
I am a former teacher. I am also a retired teacher. I was a teacher because I wanted to be a teacher. I was a retired teacher because I wanted to be a teacher again. I am also a former teacher because I wanted to be a teacher again.
I would like to thank you for your time and for your comments. I have been thinking

 

Previous                                                    Next                                                    Home