OpenAI Overview

Intro to OpenAI (Shell Script)

Introduction to OpenAI CLI and API

Prompt Engineering in AI
Tokenization in AI
Text Generation
Multimodal



1. Evolution of AI
2. TRansformers
3. LLMs Pre-training and Fine-tuning
4. Encoders
5. Bias and Fairness
6. Reinforced Learning in AI Systems
7. Etical consideration


Evolution of AI:
--------------------------------->
1. Machine Learning
2. Deep Learning
3. Natural language Processing
4. Artificial Neural Network
5. Cognitive computing
6. Computer vision


Text Generation capabilities of OpenAI
Prompt Engineering
Practical Applications
Chat Completions
Sentiment Analysis
OpenAI Assistant
Text to Speech
Speech to Text




Create apikey in platform.chatgpt.com ... and sample python code to tell a imaginary story about penguins who learns to fly?

Export OPENAI_API_KEY

export OPENAI_API_KEY="sk-proj-Uz57SbxEXMCTB9w1iOyZaM"


 (.venv) bharathkumardasaraju@11.OpenAI$ python3.12 ./penguins_story.py
Story Generated and Played Successfully!
(.venv) bharathkumardasaraju@11.OpenAI$ ls -lrth
total 632
-rw-r--r--@ 1 bharathkumardasaraju  staff   717B Mar 11 14:48 1.intro-openai.sh
-rw-r--r--@ 1 bharathkumardasaraju  staff   1.0K Mar 11 14:51 penguins_story.py
-rw-r--r--@ 1 bharathkumardasaraju  staff   266K Mar 11 15:06 tts_example.mp3
(.venv) bharathkumardasaraju@11.OpenAI$


Features of OpenAPI:
------------------------->
1. Function calling
2. Structured Outputs
3. Advanced Usage
4. Embeddings
5. Batch Processing

Process Multiple Prompts in OpenAI:
-------------------------------------------------------------------->

(.venv) bharathkumardasaraju@11.OpenAI$ python3.12 ./openai_multi_prompts.py
Prompt: Tell me a story about a brave knight
Response:
Once upon a time, in a kingdom nestled between vast forests and towering mountains, there lived a brave knight named Sir Cedric. Known throughout the realm for his unwavering courage and noble heart, Sir Cedric served as the protector of the kingdom of Eldoria. His armor, though dented and weathered from countless battles, shone brightly in the sun, a testament to his valor and dedication.

One day, news arrived of a fearsome dragon that had taken residence in the caves of Mount Draconis, threatening the peace of Eldoria. The dragon, with scales as dark as the night and eyes like burning embers, had begun to terrorize the nearby villages, burning crops and homes with its fiery breath. The people of Eldoria

================================================================================
Prompt: Generate a list of 5 business ideas
Response:
Certainly! Here are five business ideas that could be viable in today's market:

1. **Eco-Friendly Packaging Solutions**: With the increasing focus on sustainability, there's a growing demand for eco-friendly packaging. This business could offer biodegradable, recyclable, or reusable packaging options for various industries, including food, retail, and e-commerce.

2. **Remote Team Building Services**: As remote work becomes more prevalent, companies are seeking innovative ways to build team cohesion. A business that offers virtual team-building activities, workshops, and retreats can help remote teams connect and improve their collaboration skills.

3. **AI-Powered Personal Health Assistant**: Develop a mobile app that uses artificial intelligence to provide personalized health recommendations, track fitness goals, and offer dietary advice. The

================================================================================
Prompt: Explain the theory of relativity in simple terms
Response:
The theory of relativity, developed by Albert Einstein, is a fundamental concept in physics that explains how space, time, and gravity interact. It consists of two main parts: special relativity and general relativity.

**Special Relativity**:
- This part deals with objects moving at constant speeds, especially those moving very fast, close to the speed of light.
- One of its key ideas is that the laws of physics are the same for all observers, no matter how fast they are moving.
- It introduces the concept that time and space are interconnected in a four-dimensional space-time.
- A famous outcome of special relativity is that time can pass at different rates for different observers, depending on their relative speeds (time dilation). For example

================================================================================
Prompt: Write a poem about basketball
Response:
In a court where echoes of sneakers speak,
The ball dances lively, its bounce unique.
Orange globe on a journey, swift and wide,
Gliding through air, with grace it does glide.

Lines of wood, polished under the light,
Witness the battles fought every night.
Players as poets, in motion they rhyme,
Crafting a story in real-time.

The dribble's rhythm, a pulsing beat,
Mirroring hearts that refuse defeat.
Crossovers sharp, like a painter's brush,
Creating beauty in the game’s rush.

The crowd's roar, a symphony's call,
As dreams take flight with each arc of the ball.
The swish of the net

================================================================================
(.venv) bharathkumardasaraju@11.OpenAI$


Vision Capabilities of OpenAI
--------------------------------------------------------------------------------------------------->
1. vision models
2. DALL-E
3. Text-to-Image Generation
4. Introduction to CLIP
5. Practical Applications
6. Ethical Considerations


(.venv) bharathkumardasaraju@11.OpenAI$ python3.12 ./generate_ai_image.py
https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-pHeIFmfUANuUJmqTD9XwQX7V.png?st=2025-03-11T07%3A08%3A18Z&se=2025-03-11T09%3A08%3A18Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-10T23%3A19%3A06Z&ske=2025-03-11T23%3A19%3A06Z&sks=b&skv=2024-08-04&sig=lrXe4a/M2iARytfPYVSydUUBlg0l/aNznDvlLS5Gu38%3D
(.venv) bharathkumardasaraju@11.OpenAI$


(.venv) bharathkumardasaraju@11.OpenAI$ python3.12 ./generate_ai_image_personal.py
https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-ItfbCstbjok1p8XaNqMxXtFn.png?st=2025-03-11T21%3A37%3A34Z&se=2025-03-11T23%3A37%3A34Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-10T23%3A20%3A49Z&ske=2025-03-11T23%3A20%3A49Z&sks=b&skv=2024-08-04&sig=XGsN9hfQapX%2B9vt6uGjv3wZqtOALPljIniFYtcjXrZE%3D
(.venv) bharathkumardasaraju@11.OpenAI$

Penguins Story (Python Script)

A storytelling script using OpenAI

import os
from openai import OpenAI
from gtts import gTTS

# Load API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

client = OpenAI(api_key=OPENAI_API_KEY)

prompt = "Tell me a story about a Penguin who learns to fly"

def get_openai_response(prompt):
    response = client.chat.completions.create(
        model="gpt-3.5-turbo",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100
    )
    return response.choices[0].message.content.strip()

def text_to_speech(text, lang='en'):
    tts = gTTS(text=text, lang=lang, slow=False)
    tts.save("tts_example.mp3")
    os.system("afplay tts_example.mp3")

# Get AI-generated story
story = get_openai_response(prompt)

# Convert the story to speech
text_to_speech(story)

print("Story Generated and Played Successfully!")

Text-to-Speech Output (Audio)

Note

The following file is an audio MP3 generated from a text-to-speech task.

• File: tts_example.mp3

• Path: ../../../11.OpenAI/tts_example.mp3

Haiku Generator (Shell Script)

Bash script to generate a Japanese-style Haiku using OpenAI

bharathkumardasaraju@~$ curl https://api.openai.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-d '{
        "model": "gpt-4o",
        "store": true,
        "messages": [
            {"role": "user", "content": "write a haiku about ai"}
        ]
    }'

{
  "id": "chatcmpl-B9oECVqrPKaRbZ8VO1Vq04JifFQck",
  "object": "chat.completion",
  "created": 1741677728,
  "model": "gpt-4o-2024-08-06",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Silent circuits hum,  \nWisdom born from lines of code—  \nDreams in silicon.",
        "refusal": null
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 13,
    "completion_tokens": 20,
    "total_tokens": 33,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "audio_tokens": 0,
      "accepted_prediction_tokens": 0,
      "rejected_prediction_tokens": 0
    }
  },
  "service_tier": "default",
  "system_fingerprint": "fp_eb9dce56a8"
}
bharathkumardasaraju@~$


"prompt_tokens": 13 → The request consumed 13 tokens (words you sent).
"completion_tokens": 20 → The response generated 20 tokens.
"total_tokens": 33 → Total tokens spent.

OpenAI Usage Overview (Image)

Chat Completions API (Shell Script)

Script demonstrating OpenAI’s Chat Completions API

Chat completions API:
----------------------------------------->

bharathkumardasaraju@~$ curl https://api.openai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{
  "model": "gpt-4o",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "hello"
        }
      ]
    },
    {
      "role": "assistant",
      "content": [
        {
          "type": "text",
          "text": "Hello! How can I assist you today?"
        }
      ]
    }
  ],
  "response_format": {
    "type": "text"
  },
  "temperature": 1,
  "max_completion_tokens": 2048,
  "top_p": 1,
  "frequency_penalty": 0,
  "presence_penalty": 0
}'
{
  "id": "chatcmpl-B9oLoto1FKLsWsocgS8KTXpw7QlH9",
  "object": "chat.completion",
  "created": 1741678200,
  "model": "gpt-4o-2024-08-06",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! How can I assist you today?",
        "refusal": null
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 21,
    "completion_tokens": 10,
    "total_tokens": 31,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "audio_tokens": 0,
      "accepted_prediction_tokens": 0,
      "rejected_prediction_tokens": 0
    }
  },
  "service_tier": "default",
  "system_fingerprint": "fp_0d4eb8a50b"
}
bharathkumardasaraju@~$





bharathkumardasaraju@~$ curl https://api.openai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{
  "model": "gpt-4o",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "hello"
        }
      ]
    },
    {
      "role": "assistant",
      "content": [
        {
          "type": "text",
          "text": "What is the capital city of AndhraPradesh State in India?"
        }
      ]
    }
  ],
  "response_format": {
    "type": "text"
  },
  "temperature": 1,
  "max_completion_tokens": 2048,
  "top_p": 1,
  "frequency_penalty": 0,
  "presence_penalty": 0
}'
{
  "id": "chatcmpl-B9oPLqCXcOgO6USEwK54lRPQg2OEM",
  "object": "chat.completion",
  "created": 1741678419,
  "model": "gpt-4o-2024-08-06",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "The capital city of Andhra Pradesh is Amaravati. However, it is important to note that the state government functions from Amaravati, Visakhapatnam, and Kurnool as part of a decentralized governance model.",
        "refusal": null
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 25,
    "completion_tokens": 44,
    "total_tokens": 69,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "audio_tokens": 0,
      "accepted_prediction_tokens": 0,
      "rejected_prediction_tokens": 0
    }
  },
  "service_tier": "default",
  "system_fingerprint": "fp_f9f4fb6dbf"
}
bharathkumardasaraju@~$

OpenAI Model Pricing (Image)

Multi-Prompt Chat (Python Script)

Python code demonstrating multi-prompt interactions

import os
from openai import OpenAI

# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# List of prompts to process
prompts = [
    "Tell me a story about a brave knight",
    "Generate a list of 5 business ideas",
    "Explain the theory of relativity in simple terms",
    "Write a poem about basketball"
]

# Function to process a single prompt
def process_prompt(prompt):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=150,
        temperature=0.7
    )
    return response.choices[0].message.content

# Process each prompt individually
for prompt in prompts:
    print(f"Prompt: {prompt}")
    response_text = process_prompt(prompt)
    print(f"Response:\n{response_text}\n")
    print("=" * 80)  # Separator for readability

Image Generation (Python Script)

Generate AI image using DALL·E

import os
from openai import OpenAI

# Load API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# Request an AI-generated image
response = client.images.generate(
    model="dall-e-3",
    prompt="A futuristic city with flying cars and neon lights.",
    size="1792x1024",
    quality="standard",
    n=1
)

# Extract the generated image URL
image_url = response.data[0].url

# Print the image URL
print(image_url)

Generated Image (DALL·E 3)

Personalized Image Generation (Python Script)

Generate personalized image via OpenAI API

import os
from openai import OpenAI

# Load API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# Request an AI-generated image
response = client.images.generate(
    model="dall-e-3",
    prompt="Create an image of ordinary boy who studies in government degree college puttur in telugu medium and now travelled and worked in countries like Germany and singapore and working now in latest field called AI,LLM,GEnAI",
    size="1792x1024",
    quality="standard",
    n=1
)

# Extract the generated image URL
image_url = response.data[0].url

# Print the image URL
print(image_url)

Generated Image (Personalized)

Generate Poem About Recursion

Python script generating a poem about recursion

import os
from openai import OpenAI

# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# Define chat prompt
completion = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Write a haiku about recursion in programming."}
    ]
)

# Print the AI-generated response
print(completion.choices[0].message.content)

How OpenAI Works (Shell Script)

Introduction to how OpenAI works

1. Machine Learning(ML) & Artificial Intelligence(AI)
 ML - Learn from data without explicitly being programmed, identify what comes next in the data - i.e. pattern recognition
 AI - Performs complext tasks like only humans can do, such as reasoning, making decisions or solving problems ..ex: Generation, language understanding, code generation

2. LargeLanguage Models
  LLM- Its AI sysem trained in vast data-systems...
  LLMs take in a prompt input text, and use probabilities to predict the next word, or sequence of words. THis prediction is based-on patterns learned during training.
  1. Prompt Engineering
  2. RAG
  3. Fine-Tuning
  LLM Optimization - how the model needs to act
  Contect Optimization - what the model needs to know

3. Generative AI(GPT-4 & DALL-E)
   GenAI relies on sophosticated Machine Learning models such as
   generative universal networks(GUNs)
   Variational Auto Encoders(VAEs)
   GPT-3, GPT-4, DALL-E

(.venv) bharathkumardasaraju@1.OpenAI-Setup$ python3.12 ./generate_poem_about_recursion.py
A loop within loops,
Mirrored code calls itself back—
Infinite echoes.
(.venv) bharathkumardasaraju@1.OpenAI-Setup$


4. Built on Neural Networks


5. Training Models

backpropagation

Help Understanding Code

Python script to help understand a given code snippet

import os
from openai import OpenAI

# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# Sample code to explain
code_snippet = """
def factorial(n):
    if n == 0:
        return 1
    else:
        return n * factorial(n - 1)
"""

# Generate explanation for the given code
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You will be provided with a piece of code, and your task is to explain it in a concise way."},
        {"role": "user", "content": f"Explain the following Python code:\n{code_snippet}"}
    ],
    temperature=0.7,
    max_tokens=64,
    top_p=1
)

# Print the AI-generated explanation
print(response.choices[0].message.content)

Translate English to French

Script to translate English to French using OpenAI

import os
from openai import OpenAI

# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# English sentence to be translated
english_sentence = "My name is Jane. What is yours?"

# Generate translation using OpenAI
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You will be provided with a sentence in English, and your task is to translate it into French."},
        {"role": "user", "content": english_sentence}
    ],
    temperature=0.7,
    max_tokens=64,
    top_p=1
)

# Print the AI-generated translation
print("English:", english_sentence)
print("French:", response.choices[0].message.content)

Medical History Generation

Generate medical history summaries

import os
from openai import OpenAI

# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# Sample unstructured medical history data
unstructured_data = """
Patient: John Doe
Age: 45
Diagnosis: Type 2 Diabetes, Hypertension
Medications: Metformin, Lisinopril
Last Visit: 2024-03-01
Next Appointment: 2024-06-15
Doctor: Dr. Smith
"""

# Generate structured CSV response
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You will be provided with unstructured medical data, and your task is to convert it into a structured CSV format."},
        {"role": "user", "content": f"Here is the patient's medical history:\n{unstructured_data}"}
    ],
    temperature=0.7,
    max_tokens=150,
    top_p=1
)

# Extract and print the AI-generated CSV data
csv_output = response.choices[0].message.content
print("Generated CSV Data:\n")
print(csv_output)

# Save the structured data to a CSV file
csv_filename = "patient_medical_history.csv"
with open(csv_filename, "w", encoding="utf-8") as file:
    file.write(csv_output)

print(f"\nCSV file saved as: {csv_filename}")

Patient Medical History (CSV)

Note

This is a sample patient history dataset used in OpenAI prompts.

• File: patient_medical_history.csv

• Path: ../../../11.OpenAI/1.OpenAI-Setup/patient_medical_history.csv

Sentiment Analysis

Python script performing sentiment analysis with OpenAI

import os
from openai import OpenAI

# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# Sample tweet for sentiment analysis
tweet = "I loved the new Batman movie!"

# Debug: Print the tweet being analyzed
print("\n Analyzing sentiment for the tweet:\n", tweet)

# Send request to OpenAI API for sentiment classification
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You will be provided with a tweet, and your task is to classify its sentiment as positive, neutral, or negative."},
        {"role": "user", "content": tweet}
    ],
    temperature=0.7,
    max_tokens=64,
    top_p=1
)

# Extract and print sentiment classification
sentiment = response.choices[0].message.content.strip()
print("\n Sentiment Classification:\n", sentiment)

Compare Two Investments

Compare investment opportunities using OpenAI

import os
from openai import OpenAI

# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

# Investment comparison prompt
prompt = "Analyze the pros and cons of VOO vs. TGRO"

# Debug: Print the request being sent
print("\n Sending the following request to OpenAI API: ", prompt)

# Send request to OpenAI API for comparison analysis
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "user", "content": prompt}
    ],
    temperature=0.8,
    max_tokens=600,
    top_p=1
)

# Extract and print the AI-generated comparison
analysis = response.choices[0].message.content.strip()
print("\n Investment Analysis:\n", analysis)

OpenAI for Industries

Shell script showing OpenAI usage across industries

Tech Industry
  1. Code Generation
  2. Debugging assistance
  (.venv) bharathkumardasaraju@1.OpenAI-Setup$ python3.12 ./help_undrstanding_code.py
      This Python code defines a recursive function named `factorial` that calculates the factorial of a given non-negative integer `n`. Here is how it works:
      - **Base Case**: If `n` is 0, the function returns 1. This is because the factorial of 0 is defined as 1 (
  (.venv) bharathkumardasaraju@1.OpenAI-Setup$

  3. Automated Documentation
  4. NLP for customer support
  5. 24/7 Virtual assistants

Education
  1. Personalized Tutoring
  2. Simplified Explanations
  3. Generate Quizzes
  5. Translations

(.venv) bharathkumardasaraju@1.OpenAI-Setup$ python3.12 ./translate_english_to_frech.py
    English: My name is Jane. What is yours?
    French: Je m'appelle Jane. Et toi, comment t'appelles-tu ?
(.venv) bharathkumardasaraju@1.OpenAI-Setup$


Healthcare
1. Medical Journal Scanning
2. Research Paper Analysis
3. Clinical Trail Insights
4. Pattern Recognition

(.venv) bharathkumardasaraju@1.OpenAI-Setup$ python3.12 ./medical_history.py
    Generated CSV Data:
    Here is the structured CSV format for the provided medical data:
    ```csv
    Patient,Age,Diagnosis,Medications,Last Visit,Next Appointment,Doctor
    John Doe,45,"Type 2 Diabetes, Hypertension","Metformin, Lisinopril",2024-03-01,2024-06-15,Dr. Smith
    ```
    CSV file saved as: patient_medical_history.csv
(.venv) bharathkumardasaraju@1.OpenAI-Setup$


Entertainment and Media

(.venv) bharathkumardasaraju@1.OpenAI-Setup$ python3.12 ./sentiment_analysis.py
     Analyzing sentiment for the tweet:
     I loved the new Batman movie!
     Sentiment Classification:
     Positive
(.venv) bharathkumardasaraju@1.OpenAI-Setup$



Finance

1. Market Data Analysis
2. Stock Movement Prediction
3. Risk Management
4. Trend identification
5. Key Insights on market trends



(.venv) bharathkumardasaraju@1.OpenAI-Setup$ python3.12 ./comapare_two_funds.py

     Sending the following request to OpenAI API:  Analyze the pros and cons of VOO vs. TGRO
     Investment Analysis:
     When comparing investments in VOO (Vanguard S&P 500 ETF) versus TGRO (Tiger Oil and Energy, Inc.), it's crucial to consider various factors such as investment goals, risk tolerance, market position, and industry focus. Here's an analysis of the pros and cons of each:

    ### VOO (Vanguard S&P 500 ETF)

    **Pros:**

    1. **Diversification:**
       - VOO is an ETF that tracks the S&P 500 Index, which includes 500 of the largest publicly traded companies in the U.S. This provides broad market exposure and diversification across various sectors.

    2. **Stability:**
       - The ETF includes well-established and financially stable companies, which can offer relatively stable returns over the long term.

    3. **Cost-Effective:**
       - Vanguard funds are known for their low expense ratios, making VOO a cost-effective way to invest in the stock market.

    4. **Liquidity:**
       - As one of the largest ETFs, VOO is highly liquid, meaning investors can buy and sell shares easily without impacting the market price significantly.

    5. **Historical Performance:**
       - Historically, the S&P 500 has delivered solid returns over the long term, making VOO an attractive option for long-term investors.

    **Cons:**

    1. **Market Risk:**
       - Although diversified, VOO is still subject to overall market risks and can experience volatility during market downturns.

    2. **Limited Upside Potential:**
       - While VOO provides stability, it may not offer the high growth potential of individual high-risk stocks or niche sectors.

    3. **Passive Management:**
       - VOO is passively managed, meaning it doesn’t actively seek to outperform the index, which might not suit investors looking for active management strategies.

    ### TGRO (Tiger Oil and Energy, Inc.)

    **Pros:**

    1. **Potential for High Returns:**
       - As a smaller company in the oil and energy sector, TGRO could offer significant upside potential if the company performs well or if there are favorable changes in the energy market.

    2. **Industry-Specific Exposure:**
       - Investors looking to gain exposure specifically to the oil and energy industry might find TGRO appealing, especially if they believe in a bullish outlook for the sector.

    **Cons:**

    1. **High Risk:**
       - TGRO is a micro-cap stock, which typically involves higher volatility and risk compared to larger, more established companies.

    2. **Lack of Diversification:**
       - Investing in a single company like TGRO does not provide the diversification benefits that an ETF like VOO offers.

    3. **Limited Information:**
       - Smaller companies often have less analyst coverage and publicly available information, making it harder to assess the company's financial health and prospects.

    4. **Liquidity Concerns:**
       - Micro-cap stocks often have lower trading volumes, which can result in higher bid-ask spreads and difficulty buying
(.venv) bharathkumardasaraju@1.OpenAI-Setup$

OpenAI Models Overview

Overview of available OpenAI models

GPT-4  -- Generative Pre-trained Transformer-4
DALL-E
Whisper
Text to Speech(TTS)
Embedding
Moderation


Generative Pre-trained Transformer-4(GPT-4) performs wide variety of tasks.. such as
1. Human-like text generation
2. language translation
3. Summarization
4. Answering questions
5. Conversational ai

GPT-4 Model developed on the Transformer Architecture.

  Pre-training and Fine-tuning

DALL-E Model- text to image generation

TTS(Text TO Speech)

Whisper model  - is a AUtomatic Speech Recognition(ASR) Model


Embeddings Model:
  Embeddings are a way of representing words, sentences and even images as vectors in a high dimensional space.

Moderation - this model designed to detect harmful content and inappropriate text.. Toxic Language Filtering

OpenAI Python Library Example

Python script showing how to use OpenAI Python SDK

from openai import OpenAI
import os


# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

completion = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {
            "role": "user",
            "content": "Write a one-sentence bedtime story about a unicorn."
        }
    ]
)

print(completion.choices[0].message.content)

Azure OpenAI Libraries (Image)

OpenAI Libraries (Shell Script)

Installation and usage of OpenAI-related libraries

openAPI libraries

(.venv) bharathkumardasaraju@1.OpenAI-Setup$ python3.12 ./opeai_lib_python.py
Once upon a time, under a shimmering silver moon, a gentle unicorn named Luna danced through fields of stardust, weaving dreams of wonder and magic for all sleeping children.
(.venv) bharathkumardasaraju@1.OpenAI-Setup$


Azure OpenAI Libraries:

Azure OpenAI client libraty for .NET
Azure OpenAI client library for JavaScript
Azure OpenAI client library for java
Azure OpenAI client libraary for Go


OpenAI Community Libraries

ML Transition Overview

Machine Learning vs Deep Learning

AI vs Data

Evolution of AI

AI evolution timeline overview

evolution of AI ...

From rule-based systems to the Deep Learning

Rule-based systems --> Machine Learning


Deep-Learning --> Neural network and Natural language processing.


The convergence of AI and data:


AGI - Artificial General Intelligence


Transformers how they power Generative-AI

Transformers in GenAI

Pretraining vs Fine-tuning

Transformer Architecture

Transformers GenAI Script

Shell script on transformer models in GenAI

Transformers training data efficiently.

Attention mechanism in transformers


Attention score

Q - Query
K - Key
V - Value

Multi-Head attention

how attention mechanison works
--------------------------------------------->
Step1: Embedding input --> Vectors
Step2: Query, Key, Value
Step3: Attention Score
Step4: Weighted sum

Fine-tuning Visual

Pretraining & Fine-tuning Shell Script

Shell script discussing LLM model training strategies

Pre-Training and Fine-Tuning

GPT-4 pre-trained on large datasets.

Sentiment analysis, summarization and translation

Pre-training:
----------------------------->
1. Initial stage where LLMs gets trained
2. Learns general patterns, structure, relationships

Fine-Tuning:
-------------------------->
1. Pre-Trained model is adapted to perform a specific task
2. Uses smaller, task-specific dataset

Prompt Example

Prompt usage example for OpenAI

from openai import OpenAI
import os


# Load OpenAI API key from environment variable
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

if not OPENAI_API_KEY:
    raise ValueError("Missing OpenAI API Key. Set it in environment variables.")

# Initialize OpenAI client
client = OpenAI(api_key=OPENAI_API_KEY)

prompt = "You are a Cricket Player. Who's better RohitSharma or Virat Kohli?"

def prompt_engine(prompt):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user","content": prompt}],
        max_tokens=500,
        temperature=.5,
        top_p=1,
        stop=["\n"]
    )
    return response.choices[0].message.content

print(prompt_engine(prompt))

Basic Prompt Engineering Script

Bash walkthrough of prompt engineering concepts

(.venv) bharathkumardasaraju@2.OpenAI-Usage$ python3.12 ./prompt_example.py
As a cricket player, subjective opinions about players can vary greatly depending on personal preferences and criteria for "better." Both Rohit Sharma and Virat Kohli are exceptional cricketers with impressive records for India.

- **Rohit Sharma** is known for his elegant batting style, ability to play big innings, and success as an opening batsman, especially in limited-overs cricket. He has several double centuries in ODIs and has been a key player in India’s success in white-ball cricket.

- **Virat Kohli** is renowned for his aggressive batting, consistency across all formats, and as one of the best chasers in ODIs. He has numerous centuries in all formats and has captained the Indian team, leading them to significant victories.

Ultimately, "better" can depend on what aspects of cricket one values more—classical stroke play versus aggressive run chasing, or limited-overs performance versus all-format excellence. Both players have made significant contributions to the Indian cricket team, and comparing them directly may not fully capture their individual strengths and roles in the team.
(.venv) bharathkumardasaraju@2.OpenAI-Usage$


(.venv) bharathkumardasaraju@2.OpenAI-Usage$ python3.12 ./prompt_example.py
As a cricket player, it's important to appreciate the unique strengths and contributions of both Rohit Sharma and Virat Kohli, as they are both exceptional talents in the cricketing world.
(.venv) bharathkumardasaraju@2.OpenAI-Usage$

Tokenization in LLMs

Tokenization Script

Script explaining LLM tokenization

in LLM tokenization converts raw text into smaller units called tokens

LLM models can only process the tokens not the raw text

tokens standardizes the way to represnt the input in the form of vector and embeddings

Word level tokenization
Sub-word level tokenization
character level tokenization

Tokenization in AI Models

Zero-shot and Few-shot Prompting

Examples of zero/few-shot prompting

Zero-Shot Prompting - which is not pre-trained
  1. Multi-lingual translation
  2. Text Summarization
  3. Text Completion
  4. Generative question answering

Few-Shot Prompting -
  1. Sentiment Analysis
  2. Text Classification
  3. Named entity recognition
  4. Customized content generation

Grounding LLMs - Visual

RAG Architecture Diagram

Grounding LLMs Shell Script

Script showing RAG and grounding LLMs

Grounding LLMs for increased Accuracy..

RAG is part of Grounding technique

Transformer Encoder-Decoder

Encoders for LLM Prompt Processing

Encoders enabling prompt processing in LLMs

Encoders helps the model to understand how everything is relates to eachother.

Translates Tokens into numerical vectors

LLM Prompt Processing Diagram

Bias and Fairness in LLMs

Addressing bias and fairness in large language models

types of Bias

1. race
2. gender
3. socio-econmic
4. culture

Bias Mitigation
  1. Fairness Indicators
  2. Model Evaluation across groups
  3. Track model behaviour

Multimodal Inputs in GenAI

Handling multimodal inputs in generative AI

multi inputs

text
speech
images
audio
video

Multimodal uses Shared Embedding Space.


Reinforcement Learning

Fakenews and deepfake issues

Ethical frameworks and Governance

AI Code Generation - v1

Generate Python code with OpenAI API

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),
    base_url=os.environ.get("OPENAI_API_BASE")
)
prompt = "Give me a motivational quote"

response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50)
print(response.choices[0].message.content)

AI Code Generation - v3

Refined AI code generation

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),
    base_url=os.environ.get("OPENAI_API_BASE")
)

prompt = "Give me a motivational quote"
response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50)
print(response.choices[0].message.content)
response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50,
        temperature=0.7)
print(response.choices[0].message.content)
prompt = """Here are some examples of motivational quotes:
1. The only way to do great work is to love what you do.
2. Success is not the key to happiness. Happiness is the key to success.
Now, generate a new motivational quote."""

response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50,
        temperature=0.7)
print(response.choices[0].message.content)

AI Code Generation - v4

Latest version of code generation example

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),
)

prompt = "Give me a motivational quote"
response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50)
print(response.choices[0].message.content)
response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50,
        temperature=0.7)
print(response.choices[0].message.content)
prompt = """Here are some examples of motivational quotes:
1. The only way to do great work is to love what you do.
2. Success is not the key to happiness. Happiness is the key to success.
Now, generate a new motivational quote."""

response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50,
        temperature=0.7)
print(response.choices[0].message.content)

OpenAI API Key Setup Script

Script to manage OpenAI API keys

sk-proj-oQkXfSYoykvLy5n43H6lWxhFLGYnHaTGqgGKMnBF

curl https://api.openai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-proj-oQkXfSYoykvLy5n43H6lWxhFLGYnHaTGqgGKMnBF" \
  -d '{
    "model": "gpt-4o-mini",
    "store": true,
    "messages": [
      {"role": "user", "content": "write a haiku about ai"}
    ]
  }'



~/openaiproject ➜  curl https://api.openai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-proj-oQkXfSYoykvLy5n43H6lWxhFLGYnHaTGqgGKMnBF" \
  -d '{
    "model": "gpt-4o-mini",
    "store": true,
    "messages": [
      {"role": "user", "content": "write a haiku about ai"}
    ]
  }'
{
  "id": "chatcmpl-BAN2N0pmvZeaRT9bVFqrDkM72xhF4",
  "object": "chat.completion",
  "created": 1741811535,
  "model": "gpt-4o-mini-2024-07-18",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Silent circuits hum,  \nThoughts woven in code and dreams,  \nMind of silicon.",
        "refusal": null,
        "annotations": []
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 13,
    "completion_tokens": 19,
    "total_tokens": 32,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "audio_tokens": 0,
      "accepted_prediction_tokens": 0,
      "rejected_prediction_tokens": 0
    }
  },
  "service_tier": "default",
  "system_fingerprint": "fp_06737a9306"
}

~/openaiproject ➜


~/openaiproject ➜  python3 ai-content-gen.py
"Success is not the key to happiness. Happiness is the key to success. If you love what you are doing, you will be successful." – Albert Schweitzer

~/openaiproject ➜

~/openaiproject ➜  env | grep -i OPENAI_
OPENAI_API_KEY=Sk-kkAI-0b53109923bafca3fd
AZURE_OPENAI_API_KEY=Sk-kkAI-0b53109923bafca3fd
OPENAI_API_BASE=https://main.bk-ai-keys.bharathkumar.com/v1
AZURE_OPENAI_ENDPOINT=https://main.bk-ai-keys.bharathkumar.com/openai
~/openaiproject ➜

~/openaiproject ➜  python3 ai-content-gen.py
"Success is not the key to happiness. Happiness is the key to success. If you love what you are doing, you will be successful." – Albert Schweitzer
"Believe you can and you're halfway there." – Theodore Roosevelt

~/openaiproject ➜



(.venv) bharathkumardasaraju@demo_prompts$ python3.12 ./ai_code_gen4.py
"Believe you can and you're halfway there." —Theodore Roosevelt
"Success is not the key to happiness. Happiness is the key to success. If you love what you are doing, you will be successful." – Albert Schweitzer
"Believe in your journey, for every step forward, no matter how small, brings you closer to your dreams."
(.venv) bharathkumardasaraju@demo_prompts$

AI Code Generation - v2

Another example of AI-based code generation

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),
    base_url=os.environ.get("OPENAI_API_BASE")
)

prompt = "Give me a motivational quote"
response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50)
print(response.choices[0].message.content)
response = client.chat.completions.create(
        model= "gpt-4o-mini",
        messages=[{"role":"user","content":prompt}],
        max_tokens=50,
        temperature=0.7,
        stop=None)
print(response.choices[0].message.content)

OpenAI Text Generation

Story Generator

Simple story generation script

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

prompt = "write a short story about Singapore?"

def short_story_generator(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=200,
        temperature=1,
        n=3,
        stop=None
    )

    return response.choices[0].message.content

print(short_story_generator(prompt))


prompt = "How solar panel works?"
def text_gen(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=200,
        temperature=0.5,
        top_p=0.3
    )

    return response.choices[0].message.content

print(text_gen(prompt))

Demo Launcher

Demo script for launching text generation examples

Text Generation

Prompt Engineering
Practical applications
chat completions

Gnerative Pre-trained Transformer(GPT)
  1. GPT Models predicts the next token.
  2. Pre-trained on large datasets
  3. Uses prompts to generate a wide variety of outputs.



(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./generate_stoty.py
**Whispers of Merlion Park**

In the heart of Singapore, where the skyline brushed the clouds and the sea glistened like scattered diamonds, there stood the iconic Merlion, a mythical creature that blended the strength of a lion and the grace of a fish. Tourists flocked to snap photos, but on this particular evening, it was a local, Mei, who stood quietly in its shadow, contemplating life.

Mei had spent her entire life in Singapore, a vibrant tapestry of cultures woven together by history and progress. The city thrummed with energy—bustling markets in Little India, the aroma of hawker centers, and the serene beauty of Gardens by the Bay. Yet, that day, Mei found herself feeling adrift amidst the vibrant chaos.

As the sun dipped below the horizon, casting hues of orange and purple across the sky, Mei noticed an elderly man seated on a nearby bench, sketching the Merlion with frail strokes. Curiosity piqued,
(.venv) bharathkumardasaraju@3.TextGeneration$





(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./generate_stoty.py
**Title: The Forgotten Orchid**

In the heart of Singapore, where skyscrapers pierced the sky and the hustle of life never ceased, lived an old gardener named Mr. Tan. His small, unassuming garden nestled between two towering buildings was filled with an explosion of colors—bright flowers blooming under the warm sun, but the crown jewel of his collection was the rare and elusive Golden Orchid.

Mr. Tan had inherited the orchid from his mother, who had cultivated it from a seedling passed down through generations. It was said that the Golden Orchid brought prosperity and good fortune to its keeper, and Mr. Tan cherished it like a piece of history. Every morning, he tended to its fragile yellow petals, whispering secrets and dreams as if the flower could listen.

As the years rolled by, Mr. Tan noticed the world around him changing. The garden that had once been a vibrant oasis was slowly being overshadowed by the rapid development of the city. Soon, workers arrived one day with plans

Solar panels work by converting sunlight into electricity through a process called the photovoltaic effect. Here’s a step-by-step explanation of how they function:

1. **Photovoltaic Cells**: Solar panels are made up of many individual solar cells, typically made from silicon. These cells are the fundamental building blocks of solar panels.

2. **Absorption of Sunlight**: When sunlight hits the solar cells, the energy from the light is absorbed by the silicon material. This energy excites electrons in the silicon, giving them enough energy to break free from their atoms.

3. **Generation of Electron Flow**: The solar cells are designed with a positive layer (p-type) and a negative layer (n-type) of silicon. The interaction between these two layers creates an electric field. When the electrons are freed by the absorbed sunlight, they are pushed towards the n-type layer, creating a flow of electric current.

4. **Direct Current (DC) Generation**: The movement of these electrons generates
(.venv) bharathkumardasaraju@3.TextGeneration$


top-P = Top of proable tokens


temperature = Controls randomness and creativity, lower values are more determistic
top_p = top probability values(Top-P sampling)
N: number of iterations
stop: when the model should stop generating


Tokenization uses technique called Byte Pair Encoding(BPE) , it tokenizes the texts,

Tokenization Example

Demonstrating tokenization techniques

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

# Define the prompt
prompt = "Is Ronaldo better than Messi?"

def generate_text(prompt):
    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100,
        temperature=0.5,
        top_p=0.3
    )

    return response.choices[0].message.content

# Generate and print the response token by token
response_text = generate_text(prompt)
for i, token in enumerate(response_text):
    print(f"Token: {token}")

Tokenization Shell Script

Tokenization demo via shell script

How GPT Processes Tokenized inputs?

Model processes tokens through layers of the transformer architecture.
Once the input text has been tokenized, the model processes these tokens throgh layers of transformer architecture.
This involves the embedding layer, which converts tokens into dense vector embeddings. Whcih encodes semantic information about tokens.

next is attention mechanism -- how tokens related to each-other
and then next is Autoregressive process -- predict next token based on the previous one.
Positional ENcoding -
Attention layers -
Decoder Stack

(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./example_tokenization.py
Token: T
Token: h
Token: e
Token:
Token: d
Token: e
Token: b
Token: a
Token: t
Token: e
Token:
Token: o
Token: v
Token: e
Token: r
Token:
Token: w
Token: h
Token: e
Token: t
Token: h
Token: e
Token: r
Token:
Token: C
Token: r
Token: i
Token: s
Token: t
Token: i
Token: a
Token: n
Token: o
Token:
Token: R
Token: o
Token: n
Token: a
Token: l
Token: d
Token: o
Token:
Token: i
Token: s
Token:
Token: b
Token: e
Token: t
Token: t
Token: e
Token: r
Token:
Token: t
Token: h
Token: a
Token: n
Token:
Token: L
Token: i
Token: o
Token: n
Token: e
Token: l
Token:
Token: M
Token: e
Token: s
Token: s
Token: i
Token:
Token: i
Token: s
Token:
Token: o
Token: n
Token: e
Token:
Token: o
Token: f
Token:
Token: t
Token: h
Token: e
Token:
Token: m
Token: o
Token: s
Token: t
Token:
Token: d
Token: i
Token: s
Token: c
Token: u
Token: s
Token: s
Token: e
Token: d
Token:
Token: t
Token: o
Token: p
Token: i
Token: c
Token: s
Token:
Token: i
Token: n
Token:
Token: s
Token: p
Token: o
Token: r
Token: t
Token: s
Token: ,
Token:
Token: a
Token: n
Token: d
Token:
Token: o
Token: p
Token: i
Token: n
Token: i
Token: o
Token: n
Token: s
Token:
Token: v
Token: a
Token: r
Token: y
Token:
Token: w
Token: i
Token: d
Token: e
Token: l
Token: y
Token: .
Token:
Token: B
Token: o
Token: t
Token: h
Token:
Token: p
Token: l
Token: a
Token: y
Token: e
Token: r
Token: s
Token:
Token: h
Token: a
Token: v
Token: e
Token:
Token: t
Token: h
Token: e
Token: i
Token: r
Token:
Token: o
Token: w
Token: n
Token:
Token: s
Token: t
Token: r
Token: e
Token: n
Token: g
Token: t
Token: h
Token: s
Token:
Token: a
Token: n
Token: d
Token:
Token: h
Token: a
Token: v
Token: e
Token:
Token: a
Token: c
Token: h
Token: i
Token: e
Token: v
Token: e
Token: d
Token:
Token: i
Token: n
Token: c
Token: r
Token: e
Token: d
Token: i
Token: b
Token: l
Token: e
Token:
Token: s
Token: u
Token: c
Token: c
Token: e
Token: s
Token: s
Token:
Token: i
Token: n
Token:
Token: t
Token: h
Token: e
Token: i
Token: r
Token:
Token: c
Token: a
Token: r
Token: e
Token: e
Token: r
Token: s
Token: .
Token:

Token:

Token: *
Token: *
Token: C
Token: r
Token: i
Token: s
Token: t
Token: i
Token: a
Token: n
Token: o
Token:
Token: R
Token: o
Token: n
Token: a
Token: l
Token: d
Token: o
Token: :
Token: *
Token: *
Token:

Token: -
Token:
Token: K
Token: n
Token: o
Token: w
Token: n
Token:
Token: f
Token: o
Token: r
Token:
Token: h
Token: i
Token: s
Token:
Token: p
Token: h
Token: y
Token: s
Token: i
Token: c
Token: a
Token: l
Token: i
Token: t
Token: y
Token: ,
Token:
Token: a
Token: e
Token: r
Token: i
Token: a
Token: l
Token:
Token: a
Token: b
Token: i
Token: l
Token: i
Token: t
Token: y
Token: ,
Token:
Token: a
Token: n
Token: d
Token:
Token: g
Token: o
Token: a
Token: l
Token: -
Token: s
Token: c
Token: o
Token: r
Token: i
Token: n
Token: g
Token:
Token: p
Token: r
Token: o
Token: w
Token: e
Token: s
Token: s
Token: .
Token:

Token: -
Token:
Token: H
Token: a
Token: s
Token:
Token: p
Token: l
Token: a
Token: y
Token: e
Token: d
Token:
Token: i
Token: n
Token:
Token: m
Token: u
Token: l
Token: t
Token: i
Token: p
Token: l
Token: e
Token:
Token: l
Token: e
Token: a
Token: g
Token: u
Token: e
Token: s
Token:
Token: (
Token: P
Token: r
Token: e
Token: m
Token: i
Token: e
Token: r
Token:
Token: L
Token: e
Token: a
Token: g
Token: u
Token: e
Token: ,
Token:
Token: L
Token: a
Token:
Token: L
Token: i
Token: g
Token: a
Token: ,
Token:
Token: S
Token: e
Token: r
Token: i
Token: e
Token:
Token: A
Token: )
Token:
Token: a
Token: n
Token: d
Token:
Token: h
Token: a
Token: s
Token:
Token: b
Token: e
Token: e
Token: n
Token:
Token: s
Token: u
Token: c
Token: c
Token: e
Token: s
Token: s
Token: f
Token: u
Token: l
Token:
Token: i
Token: n
Token:
Token: e
Token: a
Token: c
Token: h
Token: .
Token:

Token: -
Token:
Token: H
Token: o
Token: l
Token: d
Token: s
Token:
Token: n
Token: u
Token: m
Token: e
Token: r
Token: o
Token: u
Token: s
Token:
Token: r
Token: e
Token: c
Token: o
Token: r
Token: d
Token: s
Token: ,
Token:
Token: i
Token: n
Token: c
Token: l
Token: u
Token: d
Token: i
Token: n
Token: g
Token:
Token: b
Token: e
Token: i
Token: n
Token: g
Token:
Token: t
Token: h
Token: e
Token:
Token: a
Token: l
Token: l
Token: -
Token: t
Token: i
Token: m
Token: e
Token:
Token: l
Token: e
Token: a
Token: d
Token: i
Token: n
Token: g
Token:
Token: s
Token: c
Token: o
Token: r
Token: e
Token: r
Token:
Token: i
Token: n
Token:
Token: t
Token: h
Token: e
(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./example_tokenization.py | wc -l
     516
(.venv) bharathkumardasaraju@3.TextGeneration$

Prompt Engineering - Python

Prompt engineering examples using Python

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

prompt = "tell me about solar energy?"

def short_story_generator(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100
    )

    return response.choices[0].message.content

print(short_story_generator(prompt))

print("\n")
print("----------------------------------------------------------------------------------------------------------------------------------------------------------------------")

prompt = "Explain how photovoltaic cells in solar panels convert sunlight into electricity.Include details of the photons and electrons movement"
def text_gen(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100
    )

    return response.choices[0].message.content

print(text_gen(prompt))

print("----------------------------------------------------------------------------------------------------------------------------------------------------------------------")

prompt = "Summarize photosynthesis in three sentences using simple language"
def text_gen1(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100
    )

    return response.choices[0].message.content

print(text_gen1(prompt))

print("----------------------------------------------------------------------------------------------------------------------------------------------------------------------")

prompt = "Summarize the article 'the future of renewable energy.'"
def text_gen2(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100
    )

    return response.choices[0].message.content

print(text_gen2(prompt))

Generate Blog Post

Script for blog post generation

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

def generate_blog_intro():
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": "Write a blog post introduction about the benefits of remote work for both companies and employees."}],
        max_tokens=150,
        temperature=0.7
    )
    return response.choices[0].message.content

# Generate and print the blog introduction
print(generate_blog_intro())

Prompt Engineering - Bash

Shell script for prompt engineering

bad prompt
prompt = "tell me about solar energy?"

good prompt
prompt = "Explain how photovoltaic cells in solar panels convert sunlight into electricity.Include details of the photons and electrons movement"


(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./prompt_engineering.py
Solar energy is the energy harnessed from the sun's rays. It is a renewable energy source that can be converted into electricity or heat, allowing it to be utilized for various applications, from residential and commercial power generation to heating water and powering vehicles. Here are some key points about solar energy:

### How Solar Energy Works
1. **Solar Photovoltaic (PV) Systems**: This technology uses solar panels composed of semiconductor materials (typically silicon) that generate electricity when exposed to sunlight. The photovoltaic effect occurs when photons from sunlight knock electrons loose from atoms, creating an electric current.

2. **Solar Thermal Systems**: These systems capture sunlight to heat a fluid, which can then be used for domestic hot water heating or for heating buildings. In large-scale applications, concentrated solar power (CSP) systems use mirrors or lenses to focus sunlight onto a small area, generating heat that is used to produce steam and drive a turbine for electricity generation.

### Benefits of Solar Energy
- **Renewable Source**: Solar energy is abundant and inexhaustible on a human timescale.
- **Environmental Impact**: It produces no greenhouse gas emissions during operation, reducing carbon footprints.
- **Energy Independence**: By utilizing local solar resources, countries can reduce dependence on imported fuels.
- **Job Creation**: The solar industry contributes to economic growth and job creation in installation, manufacturing, and maintenance.

### Challenges and Considerations
- **Intermittency**: Solar energy production is dependent on


----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Photovoltaic cells in solar panels convert sunlight into electricity through a process known as the photovoltaic effect. Here’s how it works step-by-step, focusing on the movement of photons and electrons:

### 1. **Absorption of Photons**
When sunlight reaches the solar panel, it is composed of particles known as photons. Each photon carries a specific amount of energy that depends on its wavelength. Photovoltaic cells are typically made from semiconductor materials, most commonly silicon. When photons hit the surface of the solar cell, their energy is absorbed by the semiconductor.

### 2. **Excitation of Electrons**
The energy absorbed from the photon excites the electrons in the silicon atoms, giving them enough energy to break free from their atomic bonds. This process creates electron-hole pairs. The 'hole' is simply the absence of an electron in the atomic structure of the silicon, leading to a positively charged vacancy.

### 3. **Formation of the Electric Field**
In a typical photovoltaic cell, there are two layers of silicon that are treated to create a p-n junction:
- The **p-type silicon layer** is doped with elements (like boron) that create holes, making it rich in positive charge carriers (holes).
- The **n-type silicon layer** is doped with elements (like phosphorus) that provide extra electrons, making it rich in negative charge carriers (electrons).

When these two layers are placed together, an electric field is formed at the junction between the two layers. This electric field plays a crucial role in the movement of the charge carriers.

### 4. **Separation of Charge Carriers**
The electric field at the p-n junction drives the free electrons (created by photon absorption) toward the n-type layer and the holes toward the p-type layer. This movement separates the positive and negative charges: electrons accumulate in the n-type layer while holes accumulate in the p-type layer.

### 5. **Creation of Electric Current**
As electrons move toward the n-type layer and holes move toward the p-type layer, a potential difference (voltage) is created across the solar cell. When the solar cell is connected to an external circuit, electrons will flow through the circuit from the n-type layer (where they are in higher concentration) to the p-type layer (where there are more holes), creating an electric current.

### 6. **Output of Electricity**
The flow of electrons through the external circuit provides electric power that can be used to power
(.venv) bharathkumardasaraju@3.TextGeneration$


Vague Prompt: Summarize photosynthesis
Specific prompt: Summarize photosynthesis in three sentences using simple language





(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./generate_blog_post.py
In today's fast-paced digital world, the traditional 9-5 office setting is rapidly undergoing a significant transformation. The emergence of remote work as a preferred mode of operation for many businesses and employees alike has seen a dramatic surge, especially due to the recent global pandemic. The concept of remote work or telecommuting is not just a fleeting trend but a revolution that offers substantial benefits for both companies and their staff. This blog post aims to delve into the numerous advantages of remote work for both parties involved, exploring how it can contribute to improved productivity, lower costs, better work-life balance, and an overall increase in job satisfaction. Read on to discover why remote work might just be the game-changer your business needs.
(.venv) bharathkumardasaraju@3.TextGeneration$

Generate Summary

Automatic summary generation

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

def summarize_article(article_text):
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": f"Summarize the following article about blockchain technology: {article_text}"}],
        max_tokens=100,
        temperature=0.3
    )
    return response.choices[0].message.content

# Example article text
article_text = "Blockchain technology is a decentralized digital ledger that records transactions across many computers to ensure security and transparency."

# Generate and print the summary
print(summarize_article(article_text))

Customer Support Assistant Generator

Customer support text generation

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

def customer_support():
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": "You are a customer support agent. A customer is asking for a refund because their product was defective. Write a professional response."}],
        max_tokens=100,
        temperature=0.35
    )
    return response.choices[0].message.content

# Generate and print the customer support response
print(customer_support())

Generate Code Comments

Generate comments for source code

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

def explain_code(code_snippet):
    """
    This function sends a code snippet to OpenAI's GPT-4 model
    and returns an explanation of the given code.
    """
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": f"Explain the following code: {code_snippet}"}],
        max_tokens=500,
        temperature=0.3
    )
    return response.choices[0].message.content

# Example code snippet to be explained
code_snippet = """
def factorial(n):
    if n == 0 or n == 1:
        return 1
    else:
        return n * factorial(n - 1)
"""

# Generate and print the explanation for the code snippet
print(explain_code(code_snippet))

Practical Applications Script

Prompt-based practical app script

1. blogposts
2. summarization






(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./generate_summary.py
The article discusses blockchain technology, describing it as a decentralized digital ledger. This technology records transactions across numerous computers, providing security and transparency. It does not go into further detail about the specific uses or development of blockchain technology.
(.venv) bharathkumardasaraju@3.TextGeneration$



(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./generate_customer_support.py
Subject: Re: Request for Refund Due to Defective Product

Dear [Customer's Name],

Thank you for reaching out to our customer service team. I'm sorry to hear about the issues you've been experiencing with your product. Our aim is to provide our customers with high-quality products and your satisfaction is our top priority.

In order to assist you further, could you please provide us with more details about the defect? This will help us understand the issue better and ensure it doesn't happen in
(.venv) bharathkumardasaraju@3.TextGeneration$




This is a recursive function in Python that calculates the factorial of a given number 'n'.

The factorial of a number is the product of all positive integers less than or equal to that number. For example, the factorial of 5 (denoted as 5!) is 1*2*3*4*5 = 120.

Here's how the function works:

- If the input number 'n' is either 0 or 1, the function returns 1. This is because the factorial of 0 and 1 is 1.

- If 'n' is greater than 1, the function returns the product of 'n' and the factorial of 'n-1'. This is done by recursively calling the same function with the argument 'n-1'.

For example, if we call factorial(5), it will return 5 * factorial(4). The factorial(4) call will return 4 * factorial(3), and so on, until factorial(1) returns 1. Then, all the returned values are multiplied together to give the final result.
(.venv) bharathkumardasaraju@3.TextGeneration$ '

Sentiment Analysis (Old API)

Sentiment detection with older OpenAI API

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

def analyze_sentiment(text):
    response = client.completions.create(
        model="gpt-4",
        prompt=f"Analyze the sentiment of the following text: '{text}'",
        max_tokens=100,
        temperature=0.3
    )

    sentiment = response.choices[0].message.content
    print(f"Input: {text}")
    print(f"Sentiment: {sentiment}")


# Example usage:
text1 = "I love the new design of the product. It's amazing!"
text2 = "The product works well, but the customer service was awful."

# Run sentiment analysis on both texts
analyze_sentiment(text1)
analyze_sentiment(text2)

Sentiment Analysis (New)

Updated sentiment analysis code

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

def analyze_sentiment(text):
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[
            {"role": "system",
             "content": "You are a sentiment analysis assistant that classifies text as 'Very Positive', 'Positive', 'Neutral', 'Negative' or 'very negative'"},
            {"role": "user", "content": f"Analyze the sentiment of the following text: '{text}'"}
        ],
        max_tokens=100,
        temperature=0.3
    )

    sentiment = response.choices[0].message.content
    print(f"Input: {text}")
    print(f"Sentiment: {sentiment}")


# Example usage:
text1 = "I love the new design of the product. It's amazing! and excellent"
text2 = "The product works well, but the customer service was awful."
text3 = "The product quality is not good and also shipping was slow too!"
text4 = "The product is huge mess and very bad!"

# Run sentiment analysis on both texts
analyze_sentiment(text1)
analyze_sentiment(text2)
analyze_sentiment(text3)
analyze_sentiment(text4)

Shell script for sentiment analysis

(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./sentiment-analysis.py
Input: I love the new design of the product. Its amazing!
Sentiment: The sentiment of the text is positive. The user expresses love for the new design of the product and describes it as amazing.
(.venv) bharathkumardasaraju@3.TextGeneration$


(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./sentiment-analysis.py
Input: I love the new design of the product. Its amazing!
Sentiment: The sentiment of the text is Positive.
Input: The product works well, but the customer service was awful.
Sentiment: Negative
(.venv) bharathkumardasaraju@3.TextGeneration$


(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./sentiment-analysis.py
Input: I love the new design of the product. Its amazing! and excellent
Sentiment: The sentiment of the text is 'Very Positive'.
Input: The product works well, but the customer service was awful.
Sentiment: 'Negative'
Input: The product quality is not good and also shipping was slow too!
Sentiment: 'Negative'
Input: The product is huge mess and very bad!
Sentiment: 'Very Negative'
(.venv) bharathkumardasaraju@3.TextGeneration$

Training Data

Sample JSONL file for fine-tuning

{"messages": [{"role": "user", "content": "I love the new design of the product. It's amazing! and excellent"}, {"role": "assistant", "content": "Very Positive"}]}
{"messages": [{"role": "user", "content": "The product works well, but the customer service was awful."}, {"role": "assistant", "content": "Mixed (Positive for product, Negative for customer service)"}]}
{"messages": [{"role": "user", "content": "The product quality is not good and also shipping was slow too!"}, {"role": "assistant", "content": "Negative"}]}
{"messages": [{"role": "user", "content": "The product is a huge mess and very bad!"}, {"role": "assistant", "content": "Very Negative"}]}

Fine-tuning Sentiment Analysis

Fine-tuning sentiment analysis using OpenAI API

import json
import time
import os
from openai import OpenAI

# Set your OpenAI API key
api_key = os.environ.get("OPENAI_API_KEY")

# Initialize OpenAI client
client = OpenAI(api_key=api_key)

# Step 1: Prepare fine-tuning dataset
training_data = [
    {"messages": [
        {"role": "user", "content": "I love the new design of the product. It's amazing! and excellent"},
        {"role": "assistant", "content": "Very Positive"}
    ]},
    {"messages": [
        {"role": "user", "content": "The product works well, but the customer service was awful."},
        {"role": "assistant", "content": "Mixed (Positive for product, Negative for customer service)"}
    ]},
    {"messages": [
        {"role": "user", "content": "The product quality is not good and also shipping was slow too!"},
        {"role": "assistant", "content": "Negative"}
    ]},
    {"messages": [
        {"role": "user", "content": "The product is a huge mess and very bad!"},
        {"role": "assistant", "content": "Very Negative"}
    ]}
]

# Save dataset to JSONL file
dataset_filename = "mydata.jsonl"
with open(dataset_filename, "w") as f:
    for entry in training_data:
        f.write(json.dumps(entry) + "\n")

print(f"Dataset saved as {dataset_filename}")

# Step 2: Upload dataset to OpenAI
file_response = client.files.create(
    file=open(dataset_filename, "rb"),
    purpose="fine-tune"
)
file_id = file_response.id
print(f"Uploaded dataset. File ID: {file_id}")

# Step 3: Start fine-tuning job
fine_tune_response = client.fine_tuning.jobs.create(
    training_file=file_id,
    model="gpt-3.5-turbo"  # Choose base model for fine-tuning
)
fine_tune_job_id = fine_tune_response.id
print(f"Started fine-tuning job. Job ID: {fine_tune_job_id}")

# Step 4: Monitor fine-tuning progress
print("Monitoring fine-tuning job...")
while True:
    fine_tune_status = client.fine_tuning.jobs.retrieve(fine_tune_job_id)
    status = fine_tune_status.status
    print(f"Fine-tuning status: {status}")

    if status in ["succeeded", "failed", "cancelled"]:
        break
    time.sleep(30)  # Wait before checking again

if status == "succeeded":
    fine_tuned_model = fine_tune_status.fine_tuned_model
    print(f"Fine-tuning completed successfully! Model ID: {fine_tuned_model}")
else:
    print(f"Fine-tuning failed with status: {status}")
    exit()

# Step 5: Use the fine-tuned model for sentiment analysis
def analyze_sentiment_fine_tuned(text):
    response = client.chat.completions.create(
        model=fine_tuned_model,
        messages=[
            {"role": "system",
             "content": "You are a sentiment analysis assistant that classifies text as 'Very Positive', 'Positive', 'Neutral', 'Negative' or 'Very Negative'."},
            {"role": "user", "content": f"Analyze the sentiment of the following text: '{text}'"}
        ],
        max_tokens=100,
        temperature=0.3
    )
    sentiment = response.choices[0].message.content
    print(f"Input: {text}")
    print(f"Sentiment: {sentiment}")

# Example usage:
texts = [
    "I love the new design of the product. It's amazing! and excellent",
    "The product works well, but the customer service was awful.",
    "The product quality is not good and also shipping was slow too!",
    "The product is a huge mess and very bad!"
]

for text in texts:
    analyze_sentiment_fine_tuned(text)

Aspect-based Sentiment Analysis

Analyzing different aspects of text sentiment

from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))

def aspect_based_sentiment(text, aspects):
    aspect_analysis = {}
    for aspect in aspects:
        print(f"\n Analyzing sentiment for aspect: '{aspect}'...")
        response = client.chat.completions.create(
            model="gpt-4",
            messages=[
                {"role": "system", "content": "You are an assistant that analyzes sentiment for specific aspects of a given text."},
                {"role": "user", "content": f"Analyze the sentiment for '{aspect}' in the following text: '{text}'. Only return 'Positive', 'Neutral', or 'Negative'."}
            ],
            max_tokens=10,
            temperature=0.3
        )
        # Debug print to see full response structure
        print(f"️ Raw Response for '{aspect}': {response}")
        # Extract sentiment response correctly
        sentiment = response.choices[0].message.content.strip()

        # Debug print to verify extracted sentiment
        print(f" Extracted Sentiment for '{aspect}': {sentiment}\n")
        aspect_analysis[aspect] = sentiment # store the result
        print("now the dictionary is this \n")
        print(aspect_analysis)
        print(aspect_analysis.items())
        print("\n")

    return aspect_analysis

# Example text
text = "The phone design is sleek and modern, but performance is lagging at times."
aspects = ["design", "performance"]

# Store the function output
sentiment_results = aspect_based_sentiment(text, aspects)

# Iterate over results and print
for aspect, sentiment in sentiment_results.items():
    print(f"Aspect: {aspect}, Sentiment: {sentiment}")

Shell script for aspect-based sentiment analysis

(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./aspect_analysis.py

 Analyzing sentiment for aspect: 'design'...
️ Raw Response for 'design': ChatCompletion(id='chatcmpl-BAkyWeVYy1SsvAqf5YuHWSBda2ijF', choices=[Choice(finish_reason='stop', index=0, logprobs=None, message=ChatCompletionMessage(content="'Positive'", refusal=None, role='assistant', audio=None, function_call=None, tool_calls=None, annotations=[]))], created=1741903552, model='gpt-4-0613', object='chat.completion', service_tier='default', system_fingerprint=None, usage=CompletionUsage(completion_tokens=4, prompt_tokens=68, total_tokens=72, completion_tokens_details=CompletionTokensDetails(accepted_prediction_tokens=0, audio_tokens=0, reasoning_tokens=0, rejected_prediction_tokens=0), prompt_tokens_details=PromptTokensDetails(audio_tokens=0, cached_tokens=0)))
 Extracted Sentiment for 'design': 'Positive'

now the dictionary is this

{'design': "'Positive'"}
dict_items([('design', "'Positive'")])



 Analyzing sentiment for aspect: 'performance'...
️ Raw Response for 'performance': ChatCompletion(id='chatcmpl-BAkyXyE8FxyqZMjYwwsOfFfw3CmAn', choices=[Choice(finish_reason='stop', index=0, logprobs=None, message=ChatCompletionMessage(content="'Negative'", refusal=None, role='assistant', audio=None, function_call=None, tool_calls=None, annotations=[]))], created=1741903553, model='gpt-4-0613', object='chat.completion', service_tier='default', system_fingerprint=None, usage=CompletionUsage(completion_tokens=4, prompt_tokens=68, total_tokens=72, completion_tokens_details=CompletionTokensDetails(accepted_prediction_tokens=0, audio_tokens=0, reasoning_tokens=0, rejected_prediction_tokens=0), prompt_tokens_details=PromptTokensDetails(audio_tokens=0, cached_tokens=0)))
 Extracted Sentiment for 'performance': 'Negative'

now the dictionary is this

{'design': "'Positive'", 'performance': "'Negative'"}
dict_items([('design', "'Positive'"), ('performance', "'Negative'")])


Aspect: design, Sentiment: 'Positive'
Aspect: performance, Sentiment: 'Negative'
(.venv) bharathkumardasaraju@3.TextGeneration$

Customer Support Assistant

Generate support responses with OpenAI

import os
from openai import OpenAI

# Read API key from environment
api_key = os.getenv("OPENAI_API_KEY")

# Initialize OpenAI client
client = OpenAI(api_key=api_key)

# Define the query
query = "Help me answer this technical question about my new snowblower."


def short_story_generator(query):
    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[
            {
                "role": "user",
                "content": f"You are a customer support assistant. A user asks: '{query}'. Respond helpfully."
            }
        ],
        max_tokens=100,
        temperature=0,
        n=1,
        stop=None,
    )

    return response.choices[0].message.content


# Generate and print the response
print(short_story_generator(query))

Shell integration with OpenAI assistants

Tools that are built to perform tasks, assist with workflows, and interact with users in natural language.

(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./customer_support_assistant.py
Of course! I’d be happy to help you with your snowblower question. Please provide me with the specific technical question you have, and I’ll do my best to assist you!
(.venv) bharathkumardasaraju@3.TextGeneration$

More Training Data

Additional JSONL training dataset

{"prompt": "Generate a confidentiality agreement clause:", "completion": "The Parties agree to keep all information confidential..."}
{"prompt": "Create a termination clause for a contract:", "completion": "This Agreement may be terminated by either party upon written notice..."}

Fine-tuning Scripts

Starting a fine-tuning job

from openai import OpenAI
client = OpenAI()

# Initialize OpenAI client
client = OpenAI()

# Upload training file
file_response = client.files.create(
    file=open("file-abc123.jsonl", "rb"),
    purpose="fine-tune"
)

print(f"Uploaded File ID: {file_response.id}")

fine_tune_response = client.fine_tuning.jobs.create(
    training_file=file_response.id,
    model="gpt-4o-2024-08-06",  # Use an available fine-tunable model
    hyperparameters={
        "n_epochs": 2  # Adjust the number of epochs as needed
    }
)

print(f"Fine-Tuning Job ID: {fine_tune_response.id}")

File used in fine-tuning job

{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "user", "content": "I fell off my bike today."}, {"role": "assistant", "content": "It's great that you're getting exercise outdoors!"}]}
{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "user", "content": "I lost my tennis match today."}, {"role": "assistant", "content": "It's ok, it happens to everyone."}, {"role": "user", "content": "But I trained so hard!"}, {"role": "assistant", "content": "It will pay off next time."}, {"role": "user", "content": "I'm going to switch to golf."}, {"role": "assistant", "content": "Golf is fun too!"}, {"role": "user", "content": "I don't even know how to play golf."}, {"role": "assistant", "content": "It's easy to learn!"}]}
{"messages": [{"role": "user", "content": "I lost my book today."}, {"role": "assistant", "content": "You can read everything on ebooks these days!"}]}
{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "assistant", "content": "You're great!"}]}
{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "user", "content": "I'm hungry."}, {"role": "assistant", "content": "Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!"}]}
{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "user", "content": "I fell off my bike today."}, {"role": "assistant", "content": "It's great that you're getting exercise outdoors!"}]}
{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "user", "content": "I lost my tennis match today."}, {"role": "assistant", "content": "It's ok, it happens to everyone."}, {"role": "user", "content": "But I trained so hard!"}, {"role": "assistant", "content": "It will pay off next time."}, {"role": "user", "content": "I'm going to switch to golf."}, {"role": "assistant", "content": "Golf is fun too!"}, {"role": "user", "content": "I don't even know how to play golf."}, {"role": "assistant", "content": "It's easy to learn!"}]}
{"messages": [{"role": "user", "content": "I lost my book today."}, {"role": "assistant", "content": "You can read everything on ebooks these days!"}]}
{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "assistant", "content": "You're great!"}]}
{"messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "user", "content": "I'm hungry."}, {"role": "assistant", "content": "Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!Eat a banana!"}]}

Using a fine-tuned model

from openai import OpenAI

client = OpenAI()

completion = client.chat.completions.create(
    model="your-fine-tuned-model-id",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {
            "role": "user",
            "content": "Write a haiku about recursion in programming."
        }
    ]
)

print(completion.choices[0].message)

Fine-tuning automation script

enhanced performance on specific tasks
Customization for specific usecases.

Prepare
Upload
Train
Evaluate


EPochs: Refers to one complete pass of the dataset.

Each epoch the model adjusts its weights

(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./fine-tuning.py
Uploaded File ID: file-2yxBP5vFjLGtTuPEvdJTuS
Fine-Tuning Job ID: ftjob-6oVluELypuux3RkM6mFVsPbQ
(.venv) bharathkumardasaraju@3.TextGeneration$





Model
ft:gpt-4o-2024-08-06:bharaths::BAnLnYzs
Status
Succeeded
Job ID
ftjob-pP1HPqxVI3QON0aTA2G82D39
Training Method
Supervised
Base model
gpt-4o-2024-08-06
Output model
ft:gpt-4o-2024-08-06:bharaths::BAnLnYzs
Created at
14 Mar 2025, 08:28
Trained tokens
32,868
Epochs
2
Batch size
1
LR multiplier
2
Seed
2137629276
Checkpoints
ft:gpt-4o-2024-08-06:bharaths::BAnLntTn:ckpt-step-10
ft:gpt-4o-2024-08-06:bharaths::BAnLnYzs
Files
Training
file-abc123.jsonl
Validation
-
Metrics
Training loss
1.3878
1
6
11
20
0
2
5.17

Messages

Metrics
08:38:09
The job has successfully completed
08:38:04
New fine-tuned model created
08:38:04
Checkpoint created at step 10
08:29:57
Fine-tuning job started
08:29:55
Files validated, moving job to queued state
08:28:48
Validating training file: file-HX6CLUfYqYk6rEgT73knLo
08:28:48
Created fine-tuning job: ftjob-pP1HPqxVI3QON0aTA2G82D39

Using Fine-tuned Model

Script to query fine-tuned model

from openai import OpenAI

client = OpenAI()

# Fine-tuned model ID
fine_tuned_model_id = "ft:gpt-4o-2024-08-06:bharaths::BAnLnYzs"

# Generate a response using the fine-tuned model
response = client.chat.completions.create(
    model=fine_tuned_model_id,
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Explain recursion in simple terms."}
    ],
    max_tokens=100,
    temperature=0.3
)

print(response.choices[0].message.content)

Embedding Models Diagram

Embedding Examples

Sample text embedding

# Embedding example

import os
from openai import OpenAI

api_key = os.getenv("OPENAI_API_KEY")
client = OpenAI(api_key=api_key)

response = client.embeddings.create(
    input="Upon Moana2 release, critics complimented that the grpahics are outstanding!",
    model="text-embedding-3-small"
)

print(response.data[0].embedding)

Additional embedding example

from openai import OpenAI
client = OpenAI()

response = client.embeddings.create(
    input="Hello",
    model="text-embedding-3-small"
)

print(response.data[0].embedding)

Embedding workflows with OpenAI

# Embedding example

Represents text as vectors in high-dimensional space.

Embeddings groups similar items together.

Its used for sematic search.

(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./embedding_example.py
[0.03250191733241081, 0.007321100682020187, -0.01696735806763172, 0.009017836302518845, 0.007610174361616373, 0.026795856654644012, 0.038233112543821335, 0.05394363030791283, 0.02785160392522812, 0.008477394469082355, -0.0020376539323478937, -0.031421031802892685, -0.02930953912436962, -0.024156490340828896, -0.0024319880176335573, 0.03182322159409523, -0.055351290851831436, -0.06399835646152496, 0.031270213425159454, -0.02978713996708393, 0.03541778773069382, 0.042330414056777954, -0.020172303542494774, 0.00022623143740929663, 0.0018821199191734195, -0.004411512985825539, -0.04160144925117493, 0.004961381200700998, 0.011889718472957611, 0.00838941615074873, 0.05942345783114433, -0.020034050568938255, 0.0377303771674633, -0.060579750686883926, -0.004684876184910536, -0.03134562075138092, -0.032677873969078064, -0.013963507488369942, -0.01725643128156662, 0.004339244682341814, 0.032853830605745316, 0.0008200889569707215, 0.013209402561187744, 0.03137075901031494, 0.013523612171411514, 0.030616654083132744, -0.024194195866584778, 0.034915052354335785, -0.017482662573456764, 0.05384308099746704, -0.03121993876993656, 0.05826716497540474, -0.041551172733306885, -0.0034280347172170877, -0.0338844433426857, -0.027977287769317627, -0.00677437474951148, 0.037906333804130554, 0.002158625051379204, -0.07586294412612915, -0.016979925334453583, 0.040771931409835815, -0.008603079244494438, 0.018928030505776405, -0.0110916243866086, -0.012870054692029953, -0.026795856654644012, -0.04197850078344345, 0.004408370703458786, -0.022711122408509254, 0.0087916050106287, 0.01992093399167061, -0.00982221495360136, -0.042179595679044724, 0.008219742216169834, -0.036825451999902725, 0.0022340354043990374, -0.03612161800265312, 0.0034971609711647034, -0.06208796054124832, -0.055954575538635254, 0.018161356449127197, -0.017608346417546272, -0.06083111837506294, 0.0019386777421459556, -0.01788485236465931, 0.01691708341240883, 0.027172910049557686, -0.07903017848730087, 0.006969185080379248, -0.011525235138833523, 0.06304316222667694, 0.023176154121756554, -0.0005211964016780257, 0.011751466430723667, 0.03453800082206726, -0.02488545887172222, 0.017671188339591026, -0.00026688238722272217, -0.00388049753382802, -0.010035878047347069, -0.037931472063064575, -0.028756529092788696, 0.01836245134472847, 0.10049702972173691, 0.00708230072632432, 0.005800322629511356, 0.04049542546272278, -0.01853840984404087, 0.0021491986699402332, -0.08928600698709488, 0.037931472063064575, -0.03622216731309891, 0.025249941274523735, 0.001900972449220717, 0.03818283975124359, 0.03695113584399223, 0.011550371535122395, -0.019757544621825218, -0.013423065654933453, -0.033180609345436096, 0.00980336219072342, -0.013121423311531544, -0.02135373465716839, 0.06430000066757202, 0.02107722871005535, -0.013724707067012787, 0.030214466154575348, -0.029183855280280113, 0.02469693124294281, -0.046126075088977814, 0.04909222200512886, -0.006717816926538944, -0.04373807832598686, -0.004797991830855608, 0.019531313329935074, 0.009922762401401997, -0.04951954632997513, 0.00033836523653008044, 0.023377249017357826, 0.0014516516821458936, 0.004986518062651157, 0.018148789182305336, -0.010827687568962574, -0.004483781289309263, -0.017181020230054855, 0.017005063593387604, -0.010702003724873066, -0.0056526437401771545, -0.011518950574100018, -0.016238389536738396, 0.005561522673815489, -0.026695309206843376, 0.004829412791877985, -0.013209402561187744, 0.012455297634005547, 0.050902072340250015, -0.013372791931033134, 0.017382115125656128, 0.0681207999587059, -0.03614675626158714, -0.006023412104696035, 0.016100136563181877, -0.008848163299262524, 0.023942826315760612, -0.01995863951742649, -0.04414026811718941, 0.020775586366653442, -0.04909222200512886, 0.0180482417345047, 0.03624730184674263, 0.017168452963232994, -0.03448772430419922, -0.04579929634928703, 0.08073948323726654, -0.010796266607940197, 0.003676260821521282, -0.04713154956698418, 0.010582603514194489, -0.004512060433626175, 0.025890931487083435, 0.03433690592646599, 0.010180414654314518, 0.019719840958714485, 0.011977697722613811, -0.008565373718738556, -0.01052604615688324, 0.03162212669849396, 0.040947891771793365, 0.048338115215301514, 0.01645205356180668, 0.036649491637945175, -0.02328926883637905, 0.0023911406751722097, -0.007365090306848288, -0.06993065029382706, 0.039489954710006714, 0.007013174705207348, -0.04185281693935394, 0.028454886749386787, 0.04826270788908005, 0.026217710226774216, 0.03262760117650032, -0.05044960975646973, -0.020310556516051292, 0.006937764119356871, -0.023993100970983505, -0.004493207670748234, 0.003965334501117468, -0.027298593893647194, 0.046000391244888306, 0.046000391244888306, -0.04363752901554108, -0.036674629896879196, -0.03209972754120827, -0.02677072025835514, -0.014830727130174637, 0.07158967852592468, -0.009841067716479301, 0.006023412104696035, 0.00982221495360136, 0.016527462750673294, 0.04336102679371834, 0.02484775334596634, 0.02563956193625927, 0.015622537583112717, 0.008892152458429337, -0.01036265678703785, -0.006830932572484016, 0.028681118041276932, 0.012153655290603638, 0.02472206950187683, -0.02058706060051918, -0.014830727130174637, 0.03463854640722275, -0.02531278319656849, 0.028580572456121445, 0.04690531641244888, -0.013473338447511196, 0.02599147893488407, 0.01400121208280325, -0.011097908951342106, 0.0006920482264831662, -0.012530707754194736, -0.02168051339685917, -0.029234129935503006, -0.008691057562828064, 0.03499046340584755, -0.0023754301946610212, -0.011921139433979988, 0.007277111057192087, 0.011933708563446999, 0.01362415961921215, -0.010306098498404026, 0.0006413818337023258, -0.025689836591482162, 0.0021853328216820955, 0.05037419870495796, -0.001355032087303698, 0.00169359368737787, -0.043763212859630585, -0.032979514449834824, -0.05575348064303398, -0.01930508203804493, -0.009991888888180256, 0.020385965704917908, -0.016979925334453583, 0.05590430274605751, 0.01755807362496853, 0.006416175048798323, -0.0055960859172046185, -0.011877150274813175, 0.013498475775122643, 0.0662606731057167, -0.0037830923683941364, -0.004977091681212187, -0.02088870294392109, 0.015144937671720982, -0.011682339943945408, -0.0049393861554563046, 0.03974132239818573, 0.003619702998548746, 0.034889914095401764, -0.009627404622733593, -0.008659636601805687, -0.009614836424589157, -0.010720856487751007, -0.032527051866054535, 0.01680396869778633, 0.019858092069625854, 0.027927014976739883, -0.013636727817356586, -0.014818158932030201, -0.017331842333078384, -0.058015793561935425, 0.027298593893647194, -0.03690085932612419, 0.018789777532219887, -0.012536992318928242, -0.02722318284213543, 0.030993707478046417, -0.04682990908622742, -0.00951428897678852, 0.004376949742436409, 0.009130951948463917, 0.020850997418165207, -0.029812276363372803, -0.015094663947820663, -0.027625372633337975, 0.05585402995347977, 0.024810047820210457, -0.05967482551932335, 0.03639812394976616, -0.009954183362424374, 0.011141898110508919, -0.02913358248770237, 0.01963186077773571, -0.007691869046539068, -0.019116556271910667, 0.0007297534612007439, 0.0014783595688641071, -0.04321020469069481, 0.025677267462015152, -0.01803567260503769, -0.04926817864179611, -0.006969185080379248, 0.057261690497398376, -0.020473945885896683, 0.033331431448459625, -0.03441231697797775, -0.05037419870495796, -0.04039487987756729, -0.0019339645514264703, -0.016690852120518684, -0.016565168276429176, 0.01560996938496828, -0.00854023639112711, 0.007415363565087318, 0.018726935610175133, 0.001985809300094843, -0.02993796020746231, -0.06580821424722672, -0.02626798301935196, -0.04275774210691452, -0.01195884495973587, -0.02659476175904274, 0.023012764751911163, 0.012536992318928242, -0.05117857828736305, 0.0012772650225088, 0.004540339112281799, -0.024156490340828896, -0.0020769303664565086, 0.019242240116000175, 0.021781060844659805, 0.012153655290603638, 0.02376686967909336, 0.04049542546272278, 0.033155474811792374, 0.05354144051671028, -0.03008878044784069, -0.011902287602424622, -0.028706256300210953, 0.026569625362753868, -0.04036974161863327, 0.004288970958441496, -0.024948300793766975, 0.06610985100269318, 0.04222986847162247, 0.008546520955860615, -0.012863771058619022, 0.03690085932612419, -0.05044960975646973, 0.03576970472931862, 0.0015576976584270597, -0.020184870809316635, 0.04939386248588562, -0.01172632910311222, 0.007691869046539068, 0.0015160648617893457, 0.00196224357932806, -0.0316723994910717, 0.01432799082249403, -0.02249746024608612, 0.006431885529309511, -0.004157002549618483, 0.038207974284887314, -0.03433690592646599, -0.02234663814306259, -0.03471395745873451, 0.03639812394976616, -0.019694702699780464, -0.03370848298072815, 0.03008878044784069, -0.015886474400758743, 0.027298593893647194, -0.044718414545059204, 0.060077015310525894, 0.01759577915072441, 0.05575348064303398, 0.033004652708768845, -0.025086551904678345, -0.0021319170482456684, -0.010871677659451962, 0.02124061807990074, -0.016703421249985695, -0.052184052765369415, -0.013762412592768669, -0.01868923008441925, -0.0004654240619856864, -0.03725277632474899, -0.061183035373687744, -0.0031939479522407055, -0.0515807680785656, 0.004282686859369278, -0.04373807832598686, -0.04698072746396065, 0.027751056477427483, -0.05434582009911537, 0.07500828802585602, 0.007170279510319233, 0.03531724214553833, -0.022887080907821655, 0.010387793183326721, 0.011154466308653355, 0.03471395745873451, 0.020499082282185555, -0.05937318503856659, 0.05711086839437485, -0.040143512189388275, -0.036322712898254395, -0.001135870348662138, 0.0510026216506958, -0.009778225794434547, 0.007943237200379372, 0.0030682638753205538, 0.034286629408597946, -0.009652541019022465, 0.03770523890852928, 0.012040539644658566, -0.00871619489043951, -0.06384754180908203, 0.020951544865965843, -0.02472206950187683, 0.008835594169795513, -0.011424687691032887, -0.008640783838927746, -0.009445162490010262, -0.009451447054743767, -0.036046210676431656, -0.003252076916396618, 0.033984988927841187, -0.05173159018158913, -0.03121993876993656, 0.003798802848905325, -0.02076301909983158, 0.0226482804864645, 0.018877755850553513, -0.0522845983505249, -0.025539016351103783, -0.0055458121933043, -0.036197029054164886, -0.02785160392522812, -0.020863566547632217, 0.02439529076218605, -0.02832920290529728, -0.021416576579213142, 0.06430000066757202, -0.004571760073304176, 0.008263731375336647, 0.058518532663583755, 0.005988848861306906, -0.008747615851461887, 0.022422049194574356, -0.04039487987756729, -0.06470219045877457, 0.04388890042901039, -0.034588273614645004, 0.03531724214553833, 0.03184835985302925, -0.03876098617911339, 0.0051090600900352, 0.005690349265933037, 0.002681785263121128, 0.010042161680758, 0.049318451434373856, 0.01933022029697895, -0.017532937228679657, -0.07455582916736603, 0.006529290694743395, -0.041576310992240906, 0.019380493089556694, 0.06560711562633514, -0.012995739467442036, -0.03182322159409523, -0.030968569219112396, -0.006749237887561321, 0.0124301603063941, -0.04067138582468033, -0.0011295861331745982, 0.026343394070863724, -0.04660367593169212, 0.03338170424103737, -0.00038176553789526224, 0.03023960255086422, 0.005960569716989994, -0.0037171081639826298, 3.292457904535695e-06, 0.032200273126363754, -0.03740359842777252, -0.008345426060259342, 0.039665911346673965, 0.04486923664808273, -0.012882623821496964, -0.029686592519283295, 0.0035757136065512896, 0.029661454260349274, -0.005275591276586056, -0.009432594291865826, 0.026066888123750687, 0.008672204799950123, -0.025564152747392654, 0.024923162534832954, -0.02342752180993557, 0.005080780945718288, -0.009501720778644085, -0.009539425373077393, -0.03325602039694786, -0.01465476956218481, -0.005853738635778427, 0.04019378498196602, -0.02689640410244465, -0.005756333004683256, 0.024898026138544083, -0.04675449803471565, 0.008169468492269516, 0.0019669567700475454, -0.006466448772698641, -0.04816215857863426, -0.00036252016434445977, 0.0471566878259182, -0.03700140863656998, -0.008854446932673454, -0.023678889498114586, -0.004669165704399347, -0.015346032567322254, -0.0257903840392828, -0.030490970239043236, -0.001488571404479444, 0.00996675156056881, 0.014139465056359768, -0.02805269882082939, 0.008282584138214588, -0.008785320445895195, 0.004072166047990322, 0.001756435725837946, -0.008012363687157631, 0.01441597007215023, 0.011525235138833523, 0.05751305818557739, 0.013108855113387108, 0.013473338447511196, 0.015320895239710808, -0.04828784242272377, 0.01931765116751194, 0.032552190124988556, 0.004760286305099726, -0.007936952635645866, 0.02805269882082939, 0.026041751727461815, -0.003531724214553833, -0.004691160283982754, -0.018714366480708122, 0.0037202502135187387, 0.014503948390483856, -0.02472206950187683, 0.015710515901446342, 0.003625987097620964, 0.04222986847162247, -0.021177776157855988, -0.025740109384059906, -0.025073984637856483, -0.020649902522563934, 0.014139465056359768, -0.015258053317666054, 0.020297987386584282, 0.009250352159142494, 0.0011869295267388225, 0.013347654603421688, 0.019594155251979828, -0.0018192777642980218, -0.030616654083132744, -0.016715990379452705, 0.037303049117326736, -0.04129980504512787, 0.014868432655930519, 0.018777208402752876, 0.020385965704917908, 0.03456313535571098, 0.002840461442247033, 0.029007896780967712, 0.014365696348249912, 0.014038917608559132, -0.019103987142443657, 0.026092026382684708, 0.019217103719711304, 0.029208991676568985, 0.007980942726135254, 0.02551387809216976, -0.004128723870962858, -0.021567396819591522, 0.005467259790748358, 0.0027556244749575853, -0.01728156767785549, -0.034286629408597946, -0.0006146739469841123, 0.040294334292411804, -0.01552199013531208, 0.011776602827012539, -0.04481896013021469, -0.010752277448773384, -0.016062432900071144, -0.0023832854349166155, -0.013536181300878525, 0.030767476186156273, 0.018274473026394844, -0.025224804878234863, -0.009124668315052986, -0.031069116666913033, -0.0033526241313666105, 0.024608952924609184, -0.014302854426205158, 0.06113276258111, 0.027801329270005226, 0.008584226481616497, -0.01258726604282856, -0.01249928679317236, 0.02641880512237549, 0.016665715724229813, -0.025765247642993927, 0.015245485119521618, -0.03139589726924896, -0.030214466154575348, 0.013397928327322006, -0.0016621726099401712, 0.018764641135931015, 0.045045193284749985, 0.03370848298072815, -0.038861531764268875, -0.013159128837287426, -0.012367318384349346, 0.005388706922531128, -0.009853635914623737, 0.007270826958119869, -0.02294992282986641, -0.031571853905916214, -0.007622742559760809, 0.01886518858373165, -0.025916067883372307, -0.036498673260211945, -0.03675004094839096, 0.04579929634928703, -0.043763212859630585, 0.012373602949082851, -0.026820993050932884, 0.008263731375336647, -0.02409364841878414, -0.01597445271909237, -0.020675038918852806, -0.020021481439471245, 0.057714153081178665, 0.00479170773178339, -0.020159734413027763, 0.053139250725507736, -0.016414348036050797, 0.00019481040362734348, -0.0430845208466053, -0.025890931487083435, -0.02946036122739315, 0.02961118146777153, -0.03431176766753197, -0.038559891283512115, -0.04992173612117767, -0.00823231041431427, -0.013360222801566124, -0.028077835217118263, 0.022233523428440094, 0.03926372155547142, 0.012254202738404274, -0.06193713843822479, -0.023980531841516495, 0.04926817864179611, 0.025375626981258392, 0.011896003037691116, 0.030993707478046417, -0.027927014976739883, -0.008257447741925716, -0.008634500205516815, -0.015949316322803497, -0.025375626981258392, -0.0024382721167057753, 0.027977287769317627, -0.023326974362134933, -0.020461376756429672, -0.0036699767224490643, -0.042330414056777954, 0.0021617671009153128, 0.005156191531568766, 0.03928885981440544, 0.04828784242272377, 0.031270213425159454, 0.010947087779641151, -0.03511614724993706, 0.00518447021022439, 0.02484775334596634, -0.0038930659648030996, -0.021014386788010597, -0.0020030909217894077, -0.012782076373696327, 0.023641185835003853, 0.015056959353387356, 0.024948300793766975, -0.01338536012917757, -0.010808835737407207, 0.013008307665586472, -0.0034374610986560583, 0.02296249009668827, -0.023012764751911163, 0.02661989815533161, 0.006994321942329407, 0.003566287225112319, -0.006629838142544031, -0.029736865311861038, -0.002607945818454027, 0.004178997594863176, 0.005784612149000168, 0.040294334292411804, -0.03169753775000572, -0.001400592504069209, 0.008414552547037601, -0.025438468903303146, 0.020235145464539528, 0.02181876450777054, -0.03499046340584755, 0.0037799503188580275, -0.010557467117905617, 0.020461376756429672, 0.030490970239043236, -0.009627404622733593, -0.01569794863462448, 0.005724912043660879, 0.009181225672364235, 0.030943432822823524, -0.006981753744184971, -0.03368334844708443, -0.00435181288048625, 0.004688018001616001, 0.005922864656895399, -0.005618080496788025, -0.024445563554763794, 0.029535770416259766, 0.021504554897546768, -0.0051059178076684475, -0.004025034140795469, 0.03815770149230957, 0.019908366724848747, 0.009489151649177074, 0.019531313329935074, 0.004537197295576334, 0.004436649847775698, 0.0031436742283403873, -0.025224804878234863, 0.011204740032553673, -0.011179603636264801, 0.01774659939110279, -0.0017344410298392177, 0.012681528925895691, -0.0020235145930200815, 0.03041555918753147, 0.046779632568359375, -0.01083397213369608, -0.022082701325416565, 0.04753373935818672, 0.02341495454311371, 0.003748529125005007, -0.017294136807322502, -0.015911610797047615, -0.028153246268630028, 0.0068812062963843346, -0.024181626737117767, -0.019694702699780464, -0.007748426869511604, 0.01172632910311222, 0.008986415341496468, -0.00459375511854887, 0.001872693537734449, 0.007195416372269392, -0.05806607007980347, -0.024948300793766975, 0.029485497623682022, -0.0029582902789115906, -0.011418403126299381, 0.005514391232281923, -0.023389816284179688, 0.003817655611783266, 0.015056959353387356, -0.01053232979029417, 0.019393062219023705, -0.03954022750258446, -0.019254809245467186, -0.004992802161723375, -0.03375875949859619, 0.007157711312174797, 0.0053918492048978806, 0.07887936383485794, -0.009646257385611534, -0.005611796397715807, 0.008075205609202385, -0.01792255789041519, 0.006692680064588785, 0.02278653346002102, -0.0077987005934119225, 0.02342752180993557, 0.025916067883372307, -0.009841067716479301, -0.021969586610794067, 0.055652935057878494, 0.03803201764822006, 0.01275065541267395, 0.04363752901554108, 0.03481450304389, -0.010412930510938168, 0.0019072566647082567, -0.0021397722885012627, 0.016690852120518684, -0.0007462495123036206, 0.03561888262629509, -0.006636122241616249, 0.008502531796693802, -0.02689640410244465, -0.04150089994072914, 0.018550977110862732, -0.012216497212648392, 0.020675038918852806, -0.036021072417497635, -0.014152033254504204, -0.004480639472603798, -0.015346032567322254, -0.03350738808512688, -0.002089498797431588, 0.00491110747680068, -0.04891626536846161, -0.0016134700272232294, -0.009017836302518845, -0.011135614477097988, -0.018601251766085625, -0.028756529092788696, -0.025916067883372307, 0.02943522296845913, -0.015924179926514626, -0.029033035039901733, 0.01679139956831932, -0.024810047820210457, -0.018412725999951363, 0.0280024241656065, 0.0031766663305461407, -0.004596896935254335, -0.007151427213102579, 0.02406851202249527, 0.022409480065107346, -0.024470699951052666, 0.014302854426205158, -0.03471395745873451, -0.005935433320701122, -0.002865598304197192, 0.00855280552059412, 0.007145143114030361, 0.006466448772698641, 0.012656391598284245, 0.055150195956230164, 0.017671188339591026, 0.043285615742206573, 0.02186903916299343, -0.042380690574645996, -0.028304066509008408, -0.013825254514813423, -0.023628616705536842, 0.0002519574190955609, -0.021718217059969902, -0.019732408225536346, -0.012229066342115402, -0.036197029054164886, -0.009344615042209625, -0.02028541825711727, 0.030038507655262947, 0.013511043973267078, 0.013033444061875343, 0.04263205826282501, 0.033809032291173935, -0.01993350312113762, -0.03054124489426613, 0.0028278930112719536, 0.009156089276075363, 0.027198046445846558, -0.01931765116751194, -0.025677267462015152, 0.036347851157188416, -0.018462998792529106, -0.008106626570224762, 0.0029394375160336494, 0.031144527718424797, 0.004502634052187204, 0.007490774150937796, -0.06002674251794815, -0.02092640846967697, 0.012153655290603638, -0.008640783838927746, -0.004153860732913017, -0.056608133018016815, -0.0355183370411396, 0.003090258687734604, 6.971541733946651e-05, 0.044567592442035675, 0.025136826559901237, 0.03008878044784069, 0.003531724214553833, 0.029887687414884567, -0.02581552043557167, -0.029058171436190605, 0.007352521643042564, -0.014076623134315014, 0.020222576335072517, 0.006441311910748482, 0.0013809544034302235, 0.002461837837472558, 0.028857076540589333, 0.027801329270005226, 0.012285623699426651, 0.023151017725467682, 0.015107233077287674, -0.011889718472957611, -0.025249941274523735, -0.04403971880674362, -0.0008444402483291924, 0.00712629035115242, -0.04310965538024902, 0.04484409838914871, -0.0014752175193279982, 0.022685986012220383, 0.05339061841368675, 0.015056959353387356, -0.005294444039463997, -0.03541778773069382, -0.0009795506484806538, -0.020863566547632217, -0.02028541825711727, -0.013284812681376934, 0.011167035438120365, -0.014755317009985447, -0.017859715968370438, -0.05120371654629707, -0.02707236260175705, 0.013435633853077888, 0.02930953912436962, -0.0038867818657308817, -0.008131762966513634, -0.017005063593387604, -0.009646257385611534, 0.01964442990720272, 0.001223849249072373, -0.00909953098744154, -0.041903089731931686, 0.004109871108084917, 0.02498600445687771, -0.010148993693292141, -0.004380092024803162, 0.01313399150967598, -0.02389255352318287, 0.03481450304389, 0.003795660799369216, 0.01552199013531208, 0.009595983661711216, -0.02915871888399124, 0.013900664635002613, 0.004254407715052366, -0.022510027512907982, 0.02961118146777153, 0.001263125566765666, -0.004606323316693306, -0.026242846623063087, -0.036197029054164886, -0.011657203547656536, -0.03041555918753147, 0.020222576335072517, 0.02200729213654995, 0.056457310914993286, -0.02298762835562229, 0.02438272163271904, 0.0010329664219170809, 0.001911969855427742, -0.0002886807487811893, 0.001029824372380972, 0.012065676972270012, -0.03308006376028061, -0.048061612993478775, -0.005435838829725981, 0.01900344155728817, 0.010588888078927994, 0.016376642510294914, 0.010148993693292141, -0.0008609362994320691, 0.017067905515432358, -0.024005668237805367, 0.01107277162373066, 0.020084323361516, -0.015773357823491096, -0.054546911269426346, -0.005162475630640984, -0.03511614724993706, -0.01645205356180668, 0.004892254713922739, -0.01945590414106846, 0.029058171436190605, -0.013737275265157223, 0.00822602678090334, -0.01020555105060339, 0.016037294641137123, -0.02440785802900791, -0.004486923571676016, -0.001318897819146514, -0.006158522330224514, -0.011135614477097988, -0.016024727374315262, -0.005351001862436533, 0.010010740719735622, 0.03350738808512688, -0.006347048562020063, 0.013900664635002613, -0.0028200377710163593, -0.010972225107252598, 0.012053107842803001, 0.018148789182305336, 0.03436204046010971, 0.0068937744945287704, 0.03310520201921463, -0.0018805487779900432, -0.00490168109536171, -0.03850961849093437, 0.003509729402139783, -0.005702917464077473, 0.004763428587466478, -0.002821608679369092, -0.006224506534636021, 0.010042161680758, -0.014214875176548958, -0.005489254370331764, 0.024294743314385414, -0.0017187305493280292, 0.010771130211651325, 0.020335692912340164, -0.04846380278468132, 0.021768491715192795, 0.026242846623063087, -0.05102775618433952, 0.002087927656248212, 0.022924786433577538, 0.023779436945915222, -0.005039933603256941, -0.0017124463338404894, 0.020373398438096046, 0.04512060433626175, -0.002278024796396494, 0.01598702184855938, -0.0035034450702369213, 0.0034594556782394648, 0.002405280014500022, -0.031747810542583466, -0.021806197240948677, 0.026971815153956413, 0.0035003030207008123, 0.03230082243680954, 0.00822602678090334, -0.008684773929417133, -0.0009638401679694653, -0.036372989416122437, 0.003767381887882948, 0.040294334292411804, -0.033155474811792374, 0.022673416882753372, 0.024508405476808548, -0.023012764751911163, -0.001019612536765635, 0.020938975736498833, -0.03431176766753197, -0.01870179921388626, 0.026066888123750687, 0.0041632866486907005, 0.042028773576021194, 0.013347654603421688, 0.015396306291222572, -0.06435027718544006, -0.009042973630130291, -0.0017721462063491344, 0.011933708563446999, 0.01868923008441925, 0.03350738808512688, -0.0015930463559925556, -0.011858297511935234, -0.041400354355573654, -0.030566381290555, 0.0068874903954565525, 0.015295758843421936, -0.007855257950723171, 0.03408553823828697, -0.012021686881780624, -0.00661098537966609, -0.0063407644629478455, -0.011053919792175293, 0.000977194169536233, 0.01583619974553585, -0.012009118683636189, -0.009432594291865826, 0.021592533215880394, 0.02469693124294281, 0.0068937744945287704, -0.012775791808962822, 0.020448807626962662, -0.01692965254187584, 0.006362759042531252, 0.05349116772413254, -0.0055898018181324005, 0.028630845248699188, 0.04326047748327255, -0.021592533215880394, -0.03199917823076248, -0.004571760073304176, -0.015182643197476864, -0.0008978560217656195, 0.04129980504512787, 0.03242650628089905, -0.02230893261730671, -0.006931480020284653, -0.013749844394624233, -0.015949316322803497, -0.0255013108253479, -0.026519352570176125, 0.006001417059451342, 0.08456028252840042, 0.012511854991316795, -0.0017187305493280292, -0.015207779593765736, 0.03313033655285835, -0.016741126775741577, -0.03881125897169113, 0.06012728810310364, 0.03486477956175804, 0.004254407715052366, 0.010595172643661499, -0.012719234451651573, 0.031094254925847054, 0.013699570670723915, -0.035065874457359314, -0.029887687414884567, 0.020348260179162025, -0.0036385555285960436, -0.003849076572805643, 0.005435838829725981, 0.0040690237656235695, 0.013636727817356586, 0.027952151373028755, 0.057714153081178665, -0.0010101861553266644, 0.009017836302518845, 0.02958604507148266, -0.012084528803825378, -0.008131762966513634, 0.023842278867959976, -0.022912217304110527, -0.011977697722613811, -0.008961278945207596, 0.014466243796050549, 0.025375626981258392, 0.0027713351882994175, 0.030591517686843872, -0.013397928327322006, 0.0006252785678952932, -0.009300625883042812, -0.0018318461952731013, -0.021441712975502014, 0.012769508175551891, 0.020800724625587463, -0.0014830727595835924, 0.0022324644960463047, -0.006183659192174673, 0.007660447619855404, -0.021806197240948677, -0.01244272943586111, 0.007654163520783186, 0.024156490340828896, 0.015911610797047615, 0.0009999743197113276, 0.02092640846967697, 0.023704027757048607, -0.023704027757048607, -0.012580981478095055, -0.008244878612458706, -0.00169359368737787, 0.009998172521591187, 0.003679402871057391, -0.030139055103063583, -0.01696735806763172, -0.006237075198441744, 0.004178997594863176, -0.02012202888727188, -0.011669771745800972, -0.01613784208893776, -0.012247918173670769, 0.008200889453291893, 0.004326676018536091, -0.010350088588893414, -0.03358279913663864, 0.020323123782873154, 0.019531313329935074, 0.01227934006601572, 0.011261298321187496, 0.024734636768698692, -0.004273260477930307, -0.020825861021876335, -0.009413741528987885, 0.03700140863656998, 0.01884005218744278, -0.01696735806763172, -0.006347048562020063, 0.005096491426229477, -0.025740109384059906, -0.026971815153956413, 0.004499491769820452, -0.010155278258025646, -0.02028541825711727, 0.04210418462753296, -0.01914169266819954, -0.01479302253574133, 0.025136826559901237, 0.010017025284469128, 0.012178792618215084, 0.013435633853077888, 0.006296774838119745, -0.02214554324746132, -0.019066283479332924, -0.020034050568938255, 0.018513273447752, 0.0007808126392774284, -0.012273055501282215, -0.015710515901446342, -0.010500908829271793, -0.0024932590313255787, -0.006460164207965136, -0.010739709250628948, -0.008841878734529018, 0.04753373935818672, -0.025689836591482162, 0.006277922540903091, 0.05085179954767227, -0.012782076373696327, 0.01598702184855938, 0.004125581588596106, -0.033180609345436096, 0.008383131586015224, 0.013372791931033134, 0.037479009479284286, -0.006711532827466726, 0.002020372310653329, -0.00336519256234169, 0.010475772432982922, 0.008012363687157631, -0.006541858892887831, -0.018287040293216705, -0.014629633165895939, 0.04698072746396065, 0.010544898919761181, -0.039691049605607986, -0.04929331690073013, 0.033959850668907166, 0.023238996043801308, 0.028404613956809044, -0.003393471473827958, -0.009633688256144524, -0.013712138868868351, 0.01564767397940159, -0.00391506077721715, 0.01163206622004509, -0.009539425373077393, 0.004697444383054972, -0.00029143007122911513, -0.008691057562828064, 0.009231499396264553, 0.033482253551483154, 0.0009096388821490109, -0.018513273447752, 0.006381611805409193, 0.012807212769985199, -0.018877755850553513, -0.03873584792017937, -0.0015357029624283314, 0.00862821564078331, -0.02042367123067379, 0.0043235342018306255, -0.025840656831860542, -0.004656597040593624, -0.004512060433626175, -0.0026440799701958895, 0.001710875192657113, -0.0037579555064439774, -0.03290410712361336, 0.02075044997036457, -0.045849572867155075, 0.005963711999356747, 0.024257037788629532, 0.009526857174932957, -0.017696326598525047, 0.00808777380734682, 0.008621931076049805, -0.02137887105345726, -0.018412725999951363, 0.01930508203804493, -0.03961563855409622, 0.006937764119356871, 0.00388049753382802, 0.006331338081508875, 0.0519326813519001, 0.03750414401292801, 0.005674638319760561, -0.02107722871005535, 0.00886073149740696, -0.021730786189436913, -0.006535574793815613, -0.009200078435242176, 0.013083717785775661, -0.03833366185426712, -0.04951954632997513, 0.021906744688749313, -0.021592533215880394, 0.035367514938116074, -0.0029268693178892136, 0.012053107842803001, 0.002952005947008729, -0.00574376480653882, 0.012618687003850937, 0.011663487181067467, -0.01568537950515747, 0.01392580196261406, 0.003827081760391593, -0.0008954994264058769, 0.01741982065141201, 0.026393666863441467, 0.04595011845231056, 0.02452097460627556, -0.008119194768369198, -0.02294992282986641, -0.005646359641104937, 0.013108855113387108, 0.001083240145817399, 0.023955395445227623, -0.01645205356180668, 0.028128108009696007, 0.027273455634713173, 0.020184870809316635, 0.01630123145878315, -0.02913358248770237, 0.04700586572289467, 0.024181626737117767, -0.035216692835092545, -0.006768090650439262, 0.0012529137311503291, 0.010230688378214836, -0.027650509029626846, -0.0029849980492144823, 0.048539210110902786, 0.011927423998713493, 0.001992093399167061, 0.001611899002455175, -0.010984793305397034, -0.028530297800898552, 0.02454611100256443, -0.0633447989821434, -0.024495836347341537, 0.036171894520521164, 0.03843420743942261, 0.00027081003645434976, -0.027977287769317627, -0.014843296259641647, -0.02704722434282303, 0.03828338533639908, 0.03564402088522911, -0.009866204112768173, 0.0015278477221727371, 0.023553205654025078, -0.016338936984539032, 0.005709201563149691, 0.02121548168361187, 0.0038427922409027815, 0.008270015940070152, -0.027273455634713173, 0.017997967079281807, -0.019506176933646202, -3.488225775072351e-05, -0.028304066509008408, 0.01677883230149746, -0.0036605503410100937, -0.009639972820878029, -0.007295963820070028, 0.003415466286242008, 0.03762982785701752, -0.013762412592768669, 0.005156191531568766, 0.013950938358902931, -0.02517453208565712, -0.003156242659315467, -0.01807337813079357, 0.04801133647561073, -0.018010536208748817, -0.031421031802892685, -7.256295066326857e-05, 0.00815690029412508, 0.024106215685606003, -0.013687001541256905, -0.032376233488321304, 0.0046471706591546535, 0.027499688789248466, -0.015320895239710808, -0.023326974362134933, 0.005731196608394384, -0.003226940054446459, -0.02599147893488407, 0.002042367123067379, -0.017067905515432358, 0.011619498021900654, 0.003569429274648428, -0.0031028268858790398, -0.011839444749057293, 0.0038867818657308817, 0.034261494874954224, 0.008911005221307278, 0.005008512642234564, 0.00041475764010101557, 0.0030635506846010685, 0.01036265678703785, 0.024734636768698692, 0.0197952501475811, 0.005599228199571371, 0.006485301069915295, 0.03873584792017937, -0.01631380058825016, 0.014918706379830837, -0.004257549997419119, -0.006868637632578611, -0.058970995247364044, -0.00404388690367341, -0.006265353877097368, -0.014767885208129883, 0.006862353533506393, -0.013435633853077888, -0.0022560302168130875, 0.0014382977969944477, 0.026846131309866905, 0.013196833431720734, 0.013636727817356586, -0.0042606922797858715, 0.002677072072401643, 0.003203374333679676, -0.009715382941067219, 0.02910844422876835, 0.033029790967702866, -0.028781665489077568, -0.0017909988528117537, -0.020838428288698196, -0.0074342163279652596, 0.017168452963232994, -0.005489254370331764, -0.02105209231376648, -0.02181876450777054, -0.019254809245467186, -0.04290856420993805, -0.008276300504803658, 0.050248514860868454, 0.03023960255086422, -0.008257447741925716, 0.03089316003024578, 0.002147627528756857, 0.006211938336491585, 0.02832920290529728, -0.008351710624992847, 0.012530707754194736, 0.050600431859493256, 0.04275774210691452, -0.008244878612458706, 0.03320574760437012, 0.011946276761591434, 0.011839444749057293, -0.019242240116000175, 0.01774659939110279, 0.03461340814828873, -0.008735046721994877, 0.0011649348307400942, -0.02596634067595005, -0.0007647093734703958, 0.005438980646431446, -0.03494018688797951, -0.003181379521265626, -0.026946676895022392, -0.0005137338885106146, 0.007220553234219551, 0.014114327728748322, -0.009256636723876, -0.0004870260017924011, -0.002980284858494997, -0.021253187209367752, -0.0061051067896187305, 0.003061979543417692, 0.014453675597906113, -0.010884245857596397, 0.008427120745182037, -0.007107437588274479, 0.007817553356289864, 0.002144485479220748, 0.0008837165660224855, -0.008653352037072182, -0.011525235138833523, -0.0031908059027045965, -0.021894175559282303, 0.0037830923683941364, -0.014780454337596893, 0.03818283975124359, 0.032527051866054535, 0.014064054004848003, 0.005957427900284529]
(.venv) bharathkumardasaraju@3.TextGeneration$

What can we do these embeddings
1. Semantic search
2. Text Clustering
3. Recommendation system
4. Anomaly detection
5. Sentiment Analysis and Classification


(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./embeddings_example2.py
[0.01926439255475998, -0.0645047277212143, -0.0016729431226849556, 0.07811243087053299, 0.021637829020619392, -0.015546008944511414, -0.015044950880110264, 0.04578095301985741, -0.0058907377533614635, -0.04527989402413368, -0.00385683448985219, -0.02060934156179428, -0.007087345235049725, -0.013119829818606377, 0.029140526428818703, 0.03359731286764145, -0.06798577308654785, 0.022244375199079514, 0.014807607047259808, 0.03554880619049072, 0.05690973252058029, 0.011616653762757778, -0.013337395153939724, 0.014899907633662224, 0.020464297384023666, 0.005528129171580076, 0.019752265885472298, 0.022297117859125137, 0.01964678056538105, -0.05240020528435707, 0.032832540571689606, -0.03375554084777832, 0.0016350341029465199, -0.006236863788217306, 0.010021176189184189, -0.01922483555972576, -0.018789706751704216, 0.027874693274497986, 0.0039985813200473785, -0.03781675547361374, 0.007232388481497765, -0.024987012147903442, 0.029140526428818703, 0.008036719635128975, 0.01460982020944357, 0.004489750601351261, -0.03449394553899765, -0.021980660036206245, 0.04156151041388512, 0.023233307525515556, 0.010060734115540981, -0.0015246033435687423, 0.02518479898571968, 0.12426258623600006, 0.03797498345375061, -0.036155350506305695, 0.04841810464859009, 0.024591440334916115, 0.011952890083193779, 0.0025217763613909483, -0.013146201148629189, 0.026397889479994774, -0.005847883876413107, 0.0026651713997125626, -0.003029427956789732, -0.02863946743309498, -0.006922523491084576, 0.014636192470788956, -0.009988212026655674, 0.0038963917177170515, -0.017444757744669914, 0.04261637106537819, 0.005013884510844946, -0.01607343927025795, -0.027821950614452362, -0.0046347943134605885, -0.046492986381053925, -0.02824389562010765, -0.0015954767586663365, -0.0021838913671672344, -0.06761656701564789, 0.020055538043379784, -0.018855635076761246, 0.0015979490708559752, -0.036445435136556625, -0.012750628404319286, -0.07336556166410446, -0.009441003203392029, 0.0025267209857702255, -0.007357653230428696, -0.018143603578209877, 0.02737363427877426, -0.014135133475065231, -0.040506649762392044, 0.00975746102631092, 0.002698135795071721, 0.005238042678683996, -0.013225316070020199, 0.09435728937387466, -0.0018591919215396047, 0.03955727443099022, -0.04158788174390793, 0.028481239452958107, -0.011089223437011242, -0.002373436465859413, 0.0502377413213253, 0.007443360518664122, 0.01251328457146883, -0.03591800481081009, 0.022152073681354523, -0.08011666685342789, -0.06983178108930588, 0.0002859661472029984, -0.008854236453771591, 0.029641585424542427, 0.0321468785405159, 0.04731050133705139, -0.07874535024166107, 0.014477962628006935, -0.05722619220614433, -0.0025382584426552057, -0.004690833389759064, 0.027610978111624718, -0.03837055712938309, -0.07531705498695374, 0.017207415774464607, -0.027136290445923805, 0.01893474906682968, -0.038766130805015564, 0.023141006007790565, 0.02998441457748413, 0.013086865656077862, -0.02853398211300373, 0.006302792578935623, -0.048365361988544464, 0.03546969220042229, -0.048391733318567276, 0.012598992325365543, -0.026780275627970695, -0.01944899372756481, 0.02750549279153347, -0.035126861184835434, -0.008570742793381214, -0.020240139216184616, -0.012361648492515087, -0.024841969832777977, -0.02263994701206684, 0.05094977095723152, -0.04382946342229843, -0.021018099039793015, 0.024195866659283638, 0.04361848905682564, -0.05108162760734558, 0.026780275627970695, 0.016442641615867615, -0.04712590202689171, 0.03335997089743614, -0.03359731286764145, 0.001400986802764237, 0.007608182728290558, 0.04451512172818184, 0.020240139216184616, -0.009038837626576424, -0.041693367063999176, -0.03829144313931465, 0.01368681713938713, 0.019607223570346832, 0.052004631608724594, -0.007304910104721785, 0.009256402961909771, -0.017418386414647102, 0.05123985931277275, -0.038080472499132156, -0.0061280811205506325, -0.01222319807857275, -0.0038238700944930315, 0.022877290844917297, 0.009328924119472504, -0.06777480244636536, -0.01520317979156971, 0.0267934612929821, 0.1044839546084404, 0.009671754203736782, -0.026674790307879448, -0.014623005874454975, -0.015440523624420166, 0.006747812032699585, -0.038344185799360275, 0.01716785691678524, 0.04462060704827309, -0.006609361618757248, 0.05279577895998955, 0.037447553128004074, 0.012420984916388988, -0.0012839632108807564, -0.0444360077381134, 0.09182561933994293, -0.006398389581590891, -0.019383065402507782, -0.0018080971203744411, 0.02924601174890995, -0.051872774958610535, 0.019172092899680138, -0.03314899653196335, -0.011972668580710888, -0.010772764682769775, 0.018011746928095818, 0.05532744154334068, 0.020371997728943825, -0.041693367063999176, 0.06972628831863403, -0.037632156163454056, -0.007384025026112795, 0.03860790282487869, -0.02319374866783619, -0.06745833903551102, 0.05822831019759178, -0.004153513815253973, 0.03739481046795845, 0.0008352354052476585, -0.02283773384988308, 0.03968913480639458, 0.06170935183763504, 0.027400005608797073, 0.02351020835340023, 0.002384974155575037, 0.0011306788073852658, -0.010159626603126526, 0.011432052589952946, -0.01871059089899063, 0.005627022590488195, 0.05859751254320145, -0.02558037079870701, -0.04092859476804733, 0.028428494930267334, 0.03971550613641739, 0.029931671917438507, -0.0013334098039194942, 0.0005278423777781427, -0.002680005505681038, -0.03771127015352249, 0.011623246595263481, 0.0008455367642454803, 0.01687777042388916, 0.027716465294361115, 0.03668278083205223, -0.003294791327789426, -0.06244775280356407, 0.0034777438268065453, 0.0021426856983453035, 0.009414631873369217, 0.028138408437371254, 0.04048027843236923, -0.02415630966424942, 0.007456546649336815, 0.02035881206393242, -0.008498221635818481, 0.011919925920665264, -0.015374594368040562, -0.04090222343802452, -0.0013704948360100389, -0.019976424053311348, 0.007265353109687567, -0.03902984410524368, -0.02190154418349266, -0.042695485055446625, -0.010542013682425022, 0.02122907154262066, -0.01794581674039364, 0.015480080619454384, 0.02515842765569687, -0.02078075520694256, 0.02908778376877308, 0.019119350239634514, -0.012341869994997978, 0.0029486652929335833, 0.005254524759948254, 0.0051424456760287285, 0.029878929257392883, 0.02821752429008484, -0.05474726855754852, -0.0009320683311671019, 0.0052017816342413425, 0.004552382975816727, -0.052162859588861465, 0.017919445410370827, 0.0024987012147903442, -0.02779557928442955, 0.026041872799396515, 0.026384703814983368, 0.030643703415989876, 0.0037282733246684074, 0.01613936759531498, -0.025936387479305267, 0.002253116574138403, 0.014280176721513271, 0.02192791737616062, -0.03219962120056152, 0.023892594501376152, -0.032779794186353683, 0.06661444902420044, 0.0028728472534567118, 0.016706356778740883, -0.004812801722437143, -0.01587565243244171, 0.05453629791736603, -0.0033030326012521982, -0.0283493809401989, 0.013594517484307289, -0.0155723812058568, -0.021624643355607986, 0.016627240926027298, 0.05975785851478577, -0.011458423919975758, -0.03027450107038021, 0.0028728472534567118, 0.018407318741083145, -0.010284891352057457, 0.015836095437407494, -0.03401925787329674, 0.008728972636163235, -0.02982618659734726, -0.000558746512979269, -0.0024987012147903442, -0.0037414589896798134, -0.011484796181321144, 0.04498980939388275, 0.0012839632108807564, -0.02766372077167034, -0.02598913013935089, 0.008959722705185413, -0.039794620126485825, 0.02145322971045971, -0.007476325146853924, 0.025079313665628433, 0.011056258343160152, -0.04351300373673439, 0.04158788174390793, -0.028402123600244522, 0.04480520635843277, -0.019765451550483704, -0.046334754675626755, -0.03897710144519806, -0.004430415108799934, -0.004094177857041359, -0.0011199653381481767, 0.021980660036206245, -0.017919445410370827, -0.020108282566070557, 0.012051783502101898, 0.023431092500686646, 0.024855155497789383, -0.012809964828193188, -0.021215885877609253, -0.032779794186353683, -0.03291165456175804, 0.061762094497680664, -0.014952650293707848, -0.016890956088900566, 0.05218923091888428, -0.03011627309024334, -0.005360011011362076, -0.013119829818606377, -0.031065646559000015, 0.03800135478377342, 0.04915650933980942, -0.048497218638658524, 0.00985635444521904, 0.01948855072259903, 0.030590960755944252, -0.020266510546207428, 0.0010499160271137953, 0.01755024492740631, 0.023075077682733536, -0.053323209285736084, 0.008320213295519352, 0.017893074080348015, -0.023391535505652428, 0.009460781700909138, -0.0035865262616425753, 0.025725414976477623, 0.04715227335691452, -0.007594997063279152, -0.012915450148284435, 0.06297518312931061, -0.004440304357558489, 0.015335037373006344, -0.04380308836698532, -0.03351819887757301, 0.019317137077450752, 0.09256402403116226, 0.010772764682769775, 0.035733405500650406, -0.0302481297403574, 0.010640907101333141, -0.02261357568204403, 0.03436208516359329, 0.0027409896720200777, -0.019567666575312614, -0.017273344099521637, 0.023299235850572586, -0.06181483715772629, 0.003988692071288824, 0.022784991189837456, -0.011247452348470688, -0.012493506073951721, -0.016759099438786507, 0.003011297667399049, 0.031804051250219345, -0.011240859515964985, 0.023984894156455994, -0.04601829871535301, 0.042115312069654465, 0.02998441457748413, -0.0019185278797522187, -0.0306964460760355, -0.014201061800122261, -0.014280176721513271, 0.006480800453573465, 0.0030557995196431875, 0.01306049432605505, 0.004690833389759064, -0.021018099039793015, 0.03496863320469856, -0.025593558326363564, 0.024235423654317856, 0.05601310357451439, -0.050844285637140274, 0.009427817538380623, -0.08591840416193008, -0.018473247066140175, 0.016693171113729477, 0.007060973905026913, 0.025013383477926254, -0.02344427816569805, -0.024987012147903442, 0.003593119326978922, 0.027690093964338303, 0.012994565069675446, 0.01993686705827713, -0.01710192859172821, -0.00851800013333559, -0.021848801523447037, -0.04702041670680046, 0.019396251067519188, -0.023905780166387558, 0.0016391546232625842, -0.01642945595085621, -0.015611938200891018, -0.013337395153939724, -0.03056458756327629, -0.009236623533070087, 0.009045430459082127, 0.006322571076452732, -0.006286310497671366, -0.0008488332387059927, 0.017049185931682587, -0.021308185532689095, 0.02319374866783619, -0.0085377786308527, 0.011207894422113895, -0.021426858380436897, 0.02966795675456524, 0.051582686603069305, 0.019277578219771385, -0.029641585424542427, -0.010509049519896507, -0.014042832888662815, -0.03612897917628288, -0.05382426828145981, 0.03837055712938309, 0.030327245593070984, 0.05060693994164467, -0.019396251067519188, -0.005379789508879185, -0.015189994126558304, 0.0028349380008876324, 0.0031744714360684156, -0.01658768393099308, -0.015229551121592522, -0.00026948394952341914, -0.016838213428854942, -0.023721178993582726, 0.05764813721179962, -0.02779557928442955, 0.044462379068136215, 0.011181523092091084, -0.005524832755327225, -0.04266911372542381, -0.03127662092447281, 0.01306049432605505, 0.027109919115900993, 0.08918847143650055, -0.014425219967961311, -0.027901064604520798, -0.02319374866783619, 0.009684939868748188, 0.008709193207323551, -0.05416709557175636, -0.062184035778045654, -0.03786949813365936, -0.009144323877990246, -0.016060253605246544, 0.014992207288742065, 0.06387181580066681, -0.045754581689834595, -0.029430612921714783, 0.015005392953753471, 0.009289367124438286, -0.033254481852054596, 0.017286529764533043, 0.04272185638546944, -0.004522715229541063, -0.04673032835125923, -0.02509249933063984, -0.009882725775241852, -0.005623726174235344, 0.020345626398921013, 0.0029915189370512962, -0.026081429794430733, 0.035443320870399475, 0.014623005874454975, -0.01261217799037695, -0.0013251687632873654, 0.0014578504487872124, -0.007660925853997469, -0.011821032501757145, -0.020411554723978043, 0.008465257473289967, 0.025461699813604355, 0.03607623651623726, -0.026951691135764122, -0.026674790307879448, -0.01431973371654749, -0.000949374632909894, 0.0027558235451579094, -0.006302792578935623, -0.039662763476371765, 0.017576616257429123, -0.026239659637212753, 0.029905300587415695, 0.013713189400732517, 0.011755104176700115, -0.013172572478652, -0.013752746395766735, 0.019383065402507782, 0.042985573410987854, -0.006728033535182476, 0.038660645484924316, 0.03177767992019653, -0.00979042612016201, -0.02583090029656887, 0.020978542044758797, -0.010621128603816032, 0.010858472436666489, -0.010673871263861656, -0.015677867457270622, -0.01483397837728262, 0.021374113857746124, 0.03734206780791283, -0.01587565243244171, 0.012249569408595562, 0.01687777042388916, 0.003695308929309249, -0.0006984331412240863, 0.03831781446933746, 0.009203659370541573, -0.014794421382248402, -0.013159386813640594, -0.013924160972237587, -0.019884124398231506, -0.008768529631197453, -0.03491589054465294, -0.024327725172042847, 0.013924160972237587, -0.03947816044092178, -0.0006151980487629771, 0.04612378403544426, 0.015677867457270622, 0.006988452281802893, -0.03291165456175804, -0.023747550323605537, -0.01761617325246334, -0.0033574237022548914, -0.028296638280153275, 0.011102409102022648, -0.012684700079262257, 0.03583889082074165, 0.013620888814330101, 0.03525871783494949, 0.030590960755944252, -0.013449473306536674, 0.019330322742462158, 0.01571742445230484, 0.018341390416026115, -0.022534461691975594, -0.019251206889748573, -0.02692531980574131, -0.021716944873332977, -0.012104526162147522, 0.001621848321519792, 0.00515233539044857, 0.015598752535879612, -0.06756382435560226, 0.04026930779218674, -0.03140847757458687, 0.010126662440598011, -0.005389678757637739, -0.017866702750325203, -0.025553999468684196, -0.017181042581796646, 0.023813480511307716, 0.014161504805088043, -0.03997921943664551, -0.030089901760220528, 0.018222717568278313, 0.006385203450918198, 0.01816997490823269, 0.004140328150242567, 0.0027475825045257807, -0.03568066284060478, -0.023615693673491478, 0.014702120795845985, 0.03620809316635132, 0.033228110522031784, 0.007621368393301964, -0.02042474038898945, 0.0036821230314671993, -0.03420385718345642, -0.0003162521752528846, 0.018249090760946274, 0.021571900695562363, 0.034441202878952026, -0.026450632140040398, -0.03860790282487869, -0.019818196073174477, 0.009223437868058681, -0.015374594368040562, 0.015888839960098267, -0.04295920208096504, 0.016732728108763695, -0.006253345869481564, 0.018249090760946274, -0.0007206841255538166, 0.02908778376877308, 0.016561312600970268, 0.011405681259930134, 0.00882786512374878, -0.0003722916590049863, -0.009770646691322327, -0.002441013464704156, -0.0368410088121891, 0.007463139481842518, -0.01613936759531498, 0.0037348661571741104, -0.004450193606317043, -0.04559635370969772, -0.011372717097401619, -0.0050633312202990055, 0.00905202329158783, 0.0011628190986812115, -0.008379549719393253, -0.020345626398921013, -0.018961120396852493, -0.002925590146332979, 0.013330802321434021, 0.012407799251377583, 0.01587565243244171, -0.012955008074641228, -0.01993686705827713, -0.0022745432797819376, 0.014504334889352322, 0.06097094714641571, -0.026978062465786934, 0.056382302194833755, 0.01984456740319729, 0.009704718366265297, -0.030801931396126747, 0.0010985385160893202, -0.026200102642178535, -0.019040236249566078, 0.011959482915699482, -0.019686337560415268, 0.007680704351514578, -0.005300675053149462, 0.028903182595968246, 0.06329163908958435, 0.012124304659664631, 0.005300675053149462, 0.02602868713438511, -0.0006514588603749871, -0.013172572478652, 0.007456546649336815, -0.0049776239320635796, 0.012579213827848434, -0.02167738787829876, 0.046202898025512695, -0.01090462226420641, 0.024235423654317856, -0.07668837159872055, 0.011504574678838253, -0.04272185638546944, -0.024367282167077065, 0.013924160972237587, -0.013554959557950497, 0.019238021224737167, -0.03902984410524368, 0.01017940603196621, 0.025501256808638573, 0.026002315804362297, -0.035416945815086365, 0.021598272025585175, -0.01594158262014389, -0.016759099438786507, 0.027452750131487846, 0.02699124813079834, -0.012625363655388355, -0.044040434062480927, 0.004921584390103817, -0.023853037506341934, -0.03467854484915733, -0.03301713988184929, -0.02734726294875145, -0.002684950130060315, 0.011260638013482094, -0.05833379551768303, -0.021598272025585175, -0.003968913108110428, 0.04483157768845558, 0.015506451949477196, -0.012974786572158337, -0.01993686705827713, 0.008689414709806442, 0.01971270889043808, 0.033834654837846756, 0.01139908842742443, -0.021268628537654877, -0.03829144313931465, -0.009777240455150604, -0.04986853897571564, 0.0042886678129434586, 0.012994565069675446, 0.017497502267360687, -0.013409916311502457, 0.02627921663224697, -0.002691542962566018, 0.01806448958814144, 0.03626083582639694, -0.01861829124391079, 0.015335037373006344, 0.021492786705493927, 0.009071801789104939, -0.027558235451579094, 0.013469252735376358, -0.0024162903428077698, -0.03127662092447281, 0.02435409650206566, -0.0008397679775953293, 0.017695287242531776, 0.018882006406784058, 0.024551883339881897, 0.028692210093140602, -0.022943219169974327, 0.004460082855075598, -0.020701641216874123, -0.005323749966919422, -0.042273543775081635, -0.007759818807244301, -0.022626761347055435, 0.0011166689218953252, 0.02821752429008484, -0.01591521129012108, 0.0013292892836034298, -0.008313620463013649, 0.013515402562916279, -0.010205777361989021, -0.0005925350124016404, 0.0014273583656176925, -0.05311223492026329, 0.02228393219411373, 0.02373436465859413, 0.028032923117280006, 0.024525510147213936, 0.0017998560797423124, -0.015822909772396088, -0.04253725707530975, -0.020728012546896935, 0.012908857315778732, 0.01119470875710249, -0.011985854245722294, 0.001976215513423085, -0.04034842178225517, 0.007951012812554836, -0.019857753068208694, 0.010311263613402843, 0.03852878510951996, 0.0029305347707122564, -0.02257401868700981, 0.016218483448028564, 0.0008364715613424778, -0.010258520022034645, -0.008399328216910362, 0.032806169241666794, -0.010957365855574608, -0.015967953950166702, -0.012809964828193188, -0.04003196209669113, 0.0071664596907794476, -0.008399328216910362, -0.006315978243947029, -0.0009304200648330152, 0.020543411374092102, -0.020543411374092102, 0.023747550323605537, 0.00847185030579567, -0.03852878510951996, 0.009137731045484543, -0.011860589496791363, 0.018816078081727028, 0.014135133475065231, 0.03881887346506119, 0.03262156620621681, -0.0034447794314473867, -0.023945337161421776, 0.004769948311150074, 0.04272185638546944, 0.024103567004203796, 0.008175170049071312, 0.004849062766879797, -0.0430910587310791, 0.013113236986100674, -0.043565746396780014, -0.027294520288705826, -0.018657848238945007, -0.0025679264217615128, 0.044910695403814316, 0.00045696893357671797, -0.005488572176545858, -0.029878929257392883, -0.026595674455165863, 0.0017652434762567282, -0.009441003203392029, 0.021242257207632065, -0.0038073877803981304, 0.010436528362333775, 0.035707034170627594, 0.005399568006396294, -0.01687777042388916, -0.02453869767487049, -0.00017790473066270351, -0.02573860064148903, -0.01298137940466404, 0.028138408437371254, -0.01125404518097639, 0.004885323345661163, 0.005178706720471382, -0.04902464896440506, 0.017339272424578667, 0.014741677790880203, -0.027742836624383926, -0.031487591564655304, 0.04393494874238968, 0.030749188736081123, 0.0010853527346625924, 0.007133495528250933, -0.04253725707530975, 0.023431092500686646, -0.020596154034137726, 0.01425380539149046, 0.01520317979156971, 0.009210252203047276, -0.0027624163776636124, -0.02225756086409092, -0.03565429151058197, 0.02035881206393242, 0.06091820448637009, -0.00627312483265996, 0.004885323345661163, -0.029878929257392883, -0.009295959956943989, 0.03462580218911171, -0.005884144920855761, -0.004252407234162092, -0.014293362386524677, 0.01980500854551792, 0.004423821810632944, 0.017629358917474747, -0.009909098036587238, -0.013330802321434021, 0.03243696689605713, -0.004624904599040747, -0.029483355581760406, 0.009157509543001652, -0.038476042449474335, 0.008788308128714561, -0.007436767686158419, -0.034573059529066086, 0.015691053122282028, -0.01203200500458479, -0.006039077416062355, -0.03243696689605713, -0.013212130405008793, -0.01716785691678524, -0.0008002107497304678, 0.012427577748894691, 0.006309385411441326, 0.013792303390800953, -0.019673151895403862, 0.01835457608103752, -0.02573860064148903, -0.005356714595109224, -0.04274822771549225, 0.01977863721549511, 0.027558235451579094, -0.001712500350549817, 0.028032923117280006, -0.015770167112350464, 0.015440523624420166, 0.00890698004513979, 0.02393215149641037, 0.002098183846101165, -0.01823590323328972, -0.019040236249566078, 0.021914731711149216, -0.015796538442373276, -0.011524353176355362, -0.01710192859172821, 0.049842167645692825, -0.0031151354778558016, 0.014135133475065231, -0.020187396556138992, 0.01112878043204546, -0.014491148293018341, 0.009361888282001019, 0.04190434142947197, -0.0015246033435687423, 0.004064510110765696, 0.029035041108727455, -0.008010348305106163, -0.000576876918785274, -0.013343987986445427, 0.010528828017413616, 0.017181042581796646, 0.02924601174890995, 0.003388739889487624, 0.054378069937229156, 0.0017800774658098817, 0.02232348918914795, 0.02016102522611618, -0.022310303524136543, 0.01335058081895113, 0.0078125623986125, 0.005748990923166275, 0.03053821623325348, -0.001327641075477004, -0.03365005552768707, 0.029008669778704643, -0.013238501735031605, 0.0011438645888119936, 0.022152073681354523, -0.025079313665628433, -0.027584606781601906, -0.035416945815086365, -0.017246972769498825, 0.009829983115196228, 0.003899688133969903, 0.008781715296208858, -0.022072959691286087, -0.009513525292277336, 0.050976142287254333, -0.026041872799396515, -0.009876132942736149, 0.01875014789402485, -0.016086624935269356, 0.038080472499132156, -0.016112996265292168, 0.017220601439476013, 0.0009815149242058396, 0.03713109716773033, -0.020147839561104774, 0.01636352576315403, 0.022020217031240463, 0.02529028430581093, -0.007397210691124201, -0.005175410304218531, 0.01232209149748087, 0.018565548583865166, 0.036287207156419754, -0.009724496863782406, 0.0056204297579824924, 0.021954288706183434, -0.007911454886198044, -0.006909337360411882, -0.004862248431891203, 0.010555199347436428, -0.0306964460760355, -0.0343884602189064, 0.010924400761723518, 0.02158508636057377, 0.019040236249566078, 0.008999280631542206, -0.032832540571689606, 0.06070723384618759, 0.02924601174890995, 0.011201301589608192, 0.017207415774464607, 0.04862907901406288, 0.02158508636057377, 0.0008030951139517128, 0.01694370061159134, 0.028138408437371254, -0.026450632140040398, 0.014820792712271214, 0.016759099438786507, 0.027821950614452362, 0.01729971542954445, 0.05780636519193649, 0.018671033903956413, -0.032595194876194, 0.019013863056898117, 0.035416945815086365, -0.0022036698646843433, -0.004802912473678589, -0.01203200500458479, 0.017049185931682587, -0.02496064081788063, 0.015361408703029156, 0.009500338695943356, 0.02779557928442955, 0.04965756833553314, -0.007575218100100756, 0.030221758410334587, 0.0032552340999245644, 0.04090222343802452, -0.001785022090189159, 0.015400965698063374, -0.0112738236784935, -0.05321772024035454, -0.015585566870868206, -0.008636672049760818, -0.028613096103072166, -0.010785950347781181, -0.011036479845643044, 0.00026371516287326813, -0.04667758569121361, 0.03235785290598869, -0.009137731045484543, -0.006856594700366259, 0.03668278083205223, -0.0006498106522485614, 0.01748431660234928, -0.00395243102684617, 0.0033063290175050497, -0.004664462059736252, 0.041825227439403534, 0.011748511344194412, -0.002061923034489155, 0.01565149612724781, -0.005472089629620314, 0.03333359956741333, -0.04237902909517288, 0.0007618896197527647, -0.042115312069654465, -0.03971550613641739, -0.0005496813100762665, -0.02924601174890995, 0.010383784770965576, 0.019092978909611702, -0.005485275760293007, 0.00964538287371397, 0.0042952606454491615, 0.01156391017138958, 0.0014611468650400639, 0.01890837773680687, 0.007555439602583647, -0.015044950880110264, 0.00948715303093195, -0.02805929444730282, -0.0028448274824768305, 0.012460541911423206, 0.004308446776121855, 0.0031942499335855246, 0.028876811265945435, -0.016534941270947456, -0.016244854778051376, 0.019633594900369644, 0.009414631873369217, -0.039214447140693665, 0.032067764550447464, 0.012249569408595562, 0.0019663262646645308, -0.0001986104907700792, -0.032595194876194, -0.004687536973506212, -0.02158508636057377, 0.012196826748549938, 0.019422622397542, 0.04699404537677765, 0.03401925787329674, 0.00012619183689821512, -0.004120549652725458, 0.010423341765999794, 0.03159307688474655, 0.013963717967271805, 0.004265592899173498, -0.018987491726875305, 0.0068697803653776646, -0.008893794380128384, -0.0036458622198551893, 0.018315019086003304, 0.04833899065852165, -0.0028217523358762264, -0.004176589194685221, -0.02422223798930645, 0.024525510147213936, -0.01526910811662674, -0.0377376414835453, -0.03436208516359329, 0.003701901761814952, -0.023971708491444588, 0.007423582021147013, -0.018591919913887978, 0.020108282566070557, 0.005666579585522413, -0.017418386414647102, 0.01828864775598049, 0.05569664388895035, -0.02107084169983864, -0.047205016016960144, 0.021242257207632065, 0.04322291538119316, 0.00325688230805099, -0.003787609050050378, -0.018301833420991898, -0.0010499160271137953, 0.044884324073791504, 0.002450902946293354, -0.03375554084777832, -0.031223876401782036, 0.026107801124453545, 0.03198865056037903, -0.036867380142211914, 0.007970791310071945, -0.007871897891163826, -0.02087305672466755, 0.003715087426826358, -0.028560353443026543, 0.11793342232704163, -0.026648418977856636, -0.04103408008813858, 0.005168817471712828, 0.007364246062934399, -0.013739560730755329, -0.0005101240240037441, 0.017062371596693993, -0.011056258343160152, -0.0005191892269067466, 0.006645622197538614, -0.011537538841366768, -0.01726015843451023, 0.012493506073951721, 0.0038007949478924274, 0.017734846100211143, 0.015559195540845394, -0.030960161238908768, -0.004526011645793915, -0.018605105578899384, 0.004344707354903221, -0.017418386414647102, -0.01349562406539917, -0.006981858983635902, 0.00866304337978363, -0.030221758410334587, -0.01032444927841425, 0.005706137046217918, 0.0354960635304451, -0.001995994243770838, 0.011359531432390213, 0.026687975972890854, 0.0341511145234108, -0.00400187773630023, 0.013245094567537308, -0.012236383743584156, -0.013462659902870655, -0.013337395153939724, 0.010746393352746964, -0.0020273104310035706, 0.013937346637248993, -0.010350820608437061, 0.017246972769498825, -0.02192791737616062, 0.022877290844917297, -0.006550025660544634, -0.0021657608449459076, 0.006573100574314594, -0.0005233098054304719, -0.002080053323879838, -0.016864584758877754, 0.04865545034408569, 0.0069488948211073875, -0.002470681443810463, 0.02428816817700863, -0.0012600640766322613, 0.005389678757637739, 0.02911415509879589, 0.04514803737401962, 0.007911454886198044, 0.004941362887620926, 0.011517760343849659, 0.03589163348078728, 0.02511887066066265, 0.017510687932372093, 0.014425219967961311, 0.0011908388696610928, 0.05295400694012642, -0.003962320275604725, 0.0033442380372434855, 0.01732608675956726, -0.015559195540845394, 0.0008982798317447305, 0.004789726808667183, -0.0030920603312551975, 0.0001290762156713754, -0.013258280232548714, 0.0175238735973835, -0.026463817805051804, 0.0048325806856155396, 0.004740280099213123, -0.007805969100445509, 0.027584606781601906, 0.018855635076761246, 0.0031019498128443956, 0.01383186038583517, 0.006177527830004692, 0.0068170372396707535, -0.022020217031240463, -0.011240859515964985, 0.010113476775586605, 0.003767830552533269, 0.004183182027190924, -0.021374113857746124, 0.023101449012756348, 0.034414831548929214, -0.04965756833553314, 0.02985255792737007, -0.000582645705435425, 0.008794900961220264, -0.02683301828801632, 0.001026840996928513, 0.017154671251773834, -0.007964198477566242, -0.01926439255475998, -0.005861069541424513, 0.035733405500650406, -0.0035074118059128523, -0.035416945815086365, -0.025870459154248238, 0.005877552088350058, -0.010106883943080902, -0.01578335277736187, 0.009183880873024464, 0.01745794527232647, -0.02882406860589981, 0.012915450148284435, -0.013403323478996754, -0.0021229072008281946, 0.0037645341362804174, 0.03201502189040184, 0.032964397221803665, 0.006012706086039543, -0.044857949018478394, 0.019976424053311348, -0.009625603444874287, 0.015282293781638145, 0.04158788174390793, 0.026081429794430733, -0.014557077549397945, -0.022653132677078247, -0.00712690269574523, -0.005112777929753065, 0.003642565803602338, 0.01185399666428566, 0.021598272025585175, 0.006058855913579464, -0.018671033903956413, 0.00861689355224371, -0.0072389813140034676, -0.013475845567882061, -0.03143484890460968, 0.006309385411441326, 0.02509249933063984, 0.008260877802968025, 0.01748431660234928, 0.02500019781291485, -0.004736983682960272, -0.008320213295519352, -0.02141367271542549, -0.02853398211300373, -0.024050824344158173, 0.025171613320708275, -0.016706356778740883, 0.019079793244600296, 0.0014941112603992224, -0.01687777042388916, 0.0061214882880449295, 0.018077675253152847, 0.00426888931542635, -0.023457463830709457, -0.02138729952275753, -0.025527628138661385, -0.01636352576315403, 0.00541934696957469, 0.02402445301413536, -0.0004013415309600532, 0.011722139082849026, -0.0005562742007896304, 0.014926278963685036, 0.040084704756736755, -0.03201502189040184, -0.0019861049950122833, 0.011511167511343956, 0.038027726113796234, 0.009309145621955395, -0.011385902762413025, 0.0061874170787632465, 0.014214247465133667, 0.009876132942736149, -0.036761894822120667, -0.015625124797225, 0.012605585157871246, -0.003316218266263604, -0.04409317672252655, 0.03649818152189255, 0.00405132444575429, -0.018024932593107224, -0.013805489055812359, -0.020622527226805687, -0.004380968399345875, -0.006988452281802893, 0.040374793112277985, 0.002704728627577424, -0.013179165311157703, -0.010687056928873062, -0.030775560066103935, -0.03968913480639458, 0.024050824344158173, -0.006395092699676752, -0.023958522826433182, 0.0036986053455621004, 0.004354596603661776, 0.029193269088864326, 0.04738961532711983, -0.010713429190218449, -0.011385902762413025, 0.004067806527018547, 0.008544371463358402, -0.0032783092465251684, 0.008465257473289967, -0.0038963917177170515, -0.004947955720126629, 0.002445958321914077, 0.05166180059313774, 0.03190953657031059, 0.0035173010546714067, 0.04628201201558113, 0.011471609584987164, 0.015440523624420166, 0.0026668196078389883, -0.02293003350496292, -0.006137970834970474, -0.02927238494157791, 0.003293143119663, 0.01884244941174984, -0.02538258582353592, -0.02412993833422661, -0.01890837773680687, 0.023246493190526962, 0.03860790282487869, -0.007291724439710379, 0.009394853375852108, 0.032806169241666794, 0.0006823629955761135, 0.02580452896654606, -0.056962478905916214, -9.580277401255444e-05, -0.013752746395766735, 0.012440763413906097, -0.0028744954615831375, -0.012922042980790138, -0.020398369058966637, -0.026740718632936478, -0.013528588227927685, 0.030063530430197716, 0.047653332352638245, -0.009124545380473137, 0.011695767752826214, 0.016917327418923378, 0.01164302509278059, 0.02515842765569687, -0.013818674720823765, -0.021268628537654877, -0.01723378710448742, -0.030459102243185043, -0.015137250535190105, 0.00300305662676692, 0.029298756271600723, 0.02045111171901226, -0.015229551121592522, 0.013258280232548714, 0.026081429794430733, -0.012809964828193188, 0.015176808461546898, -0.012190233916044235, 2.9152888600947335e-05, 0.029905300587415695, -0.02518479898571968, 0.0068038515746593475, 0.002643744694069028, -0.004107363987714052, -0.01784033142030239, 0.00897290837019682, 0.00670825457200408, 0.033096253871917725, -0.0010482678189873695, 0.009401446208357811, 0.002895922167226672, -0.010713429190218449, 0.0011265582870692015, 0.025211170315742493, -0.017642544582486153, 0.013502216897904873, 0.007977384142577648, 0.010284891352057457, 0.024169495329260826, 0.00549186859279871, 0.002511887112632394, -0.02344427816569805, 0.02953610010445118, 0.026661604642868042, 0.005439125467091799, 0.009698125533759594, 0.012974786572158337, 0.030617332085967064, -0.0006452780799008906, -0.011155151762068272, -0.00839273538440466, 0.029008669778704643, -0.00011032771726604551, -0.006474207621067762, 0.030485473573207855, -0.017761217430233955, 0.001647395663894713, 0.04390857741236687, -0.017734846100211143, 0.024274980649352074, -0.0020503855776041746, 0.0008999280398711562, -0.024340910837054253, -0.005630319006741047, -0.013291244395077229, -0.01240120641887188, -0.024037638679146767, 0.01944899372756481, -0.02422223798930645, 0.026134174317121506, 0.016020696610212326, 0.04564909636974335, 0.027479121461510658, 0.01383186038583517, 0.002829993376508355, -0.016086624935269356, 0.009058616124093533, 0.013779117725789547, -0.006177527830004692, -0.018631476908922195, 0.06508490443229675, -0.002744286088272929, -0.0028398828580975533, 0.003942541778087616, 0.007825748063623905, 0.00456556910648942, 0.029615214094519615, -0.004674351308494806, -0.005079813301563263, -0.026173731312155724, 0.032542452216148376, -0.027426378801465034, -0.00905202329158783, -0.01117493025958538, -0.0035238938871771097, 0.020029166713356972, -0.03786949813365936, 0.00021097213902976364, -0.0007206841255538166, -0.017708472907543182, 0.020015981048345566, 0.057595394551754, -0.011148558929562569, 0.002853068523108959, 0.014662563800811768, -0.004971031099557877, -0.013133015483617783, 0.0010729910572990775, 0.021848801523447037, -0.04873456433415413, 0.01331761572510004, -0.009816797450184822, -0.0005686358199454844, -0.03270068019628525, -0.012862707488238811, 0.011972668580710888, -0.009493745863437653, 0.044277776032686234, -0.027268148958683014, 0.008920165710151196, -0.012770406901836395, 0.013442880474030972, 0.010456306859850883, -0.019475365057587624, -0.01061453577131033, -0.04918288066983223, 0.005066627636551857, -0.03568066284060478, 0.023365164175629616, 0.01267151441425085, -0.004259000066667795, 0.006164342164993286, 0.005376493092626333, -0.030590960755944252, 0.0068038515746593475, 0.0009510228410363197, 0.017563430592417717, 0.01794581674039364, -0.0020998320542275906, -0.014715306460857391, -0.022890476509928703]
(.venv) bharathkumardasaraju@3.TextGeneration$


Embeddings measures the relatedness of text strengths

Text-to-Speech

Generate speech from text

import os
from gtts import gTTS
from openai import OpenAI

# Initialize OpenAI client
client = OpenAI()

# Define the prompt
prompt = "Tell me a story about a brave knight that loves basketball"

# Function to convert text to speech
def text_to_speech(text, lang='en'):
    tts = gTTS(text=text, lang=lang, slow=False)
    tts.save("tts_example.mp3")
    os.system("afplay tts_example.mp3")

# Function to generate text using OpenAI's API
def generate_text(prompt):
    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        temperature=0.8,
        max_tokens=150
    )
    return response.choices[0].message.content

# Function to generate text and convert it to speech
def gen_and_speak(prompt):
    text = generate_text(prompt)
    print("Generated Text:\n", text)
    text_to_speech(text)

# Execute the function
print(gen_and_speak(prompt))

Speech-to-Text

Convert speech to text

from openai import OpenAI

client = OpenAI()

audio_file = open("./tts_example.mp3", "rb")

transcribe = client.audio.transcriptions.create(
    model="whisper-1",
    file=audio_file
)

print(transcribe.text)

TTS/STT integration script

Text to speech

Speech to text

GTTS - Google Text To Speech library

(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./text_to_speech.py
Generated Text:
 Once upon a time in the kingdom of Eldoria, there lived a brave knight named Sir Cedric. Sir Cedric was not just any knight; he was renowned for his courage in battle, his unwavering sense of justice, and his unique passion for basketball. While most knights spent their days practicing sword fighting and jousting, Cedric could often be found dribbling a leather ball down the castle courtyard, weaving through imaginary defenders, and shooting hoops with great flair.

In Eldoria, basketball was a new and exciting sport, introduced by traveling merchants from distant lands. The game quickly captured the hearts of the villagers, and Cedric became a local hero for championing it. He organized tournaments, taught children how to play, and even fashioned a
None
(.venv) bharathkumardasaraju@3.TextGeneration$


(.venv) bharathkumardasaraju@3.TextGeneration$ python3.12 ./speech_to_text.py
Once upon a time in the kingdom of Eldoria, there lived a brave knight named Sir Cedric. Sir Cedric was not just any knight. He was renowned for his courage in battle, his unwavering sense of justice, and his unique passion for basketball. While most knights spent their days practicing sword fighting and jousting, Cedric could often be found dribbling a leather ball down the castle courtyard, weaving through imaginary defenders, and shooting hoops with great flair. In Eldoria, basketball was a new and exciting sport. Introduced by traveling merchants from distant lands, the game quickly captured the hearts of the villagers, and Cedric became a local hero for championing it. He organized tournaments, taught children how to play, and even fashioned a
(.venv) bharathkumardasaraju@3.TextGeneration$

Text Generation - Demo Scripts

Sample dataset for demo use

Person ID,Gender,Age,Occupation,Sleep Duration,Quality of Sleep,Physical Activity Level,Stress Level,BMI Category,Blood Pressure,Heart Rate,Daily Steps,Sleep Disorder
1,Male,27,Software Engineer,6.1,6,42,6,Overweight,126/83,77,4200,None
2,Male,28,Doctor,6.2,6,60,8,Normal,125/80,75,10000,None
3,Male,28,Doctor,6.2,6,60,8,Normal,125/80,75,10000,None
4,Male,28,Sales Representative,5.9,4,30,8,Obese,140/90,85,3000,Sleep Apnea
5,Male,28,Sales Representative,5.9,4,30,8,Obese,140/90,85,3000,Sleep Apnea
6,Male,28,Software Engineer,5.9,4,30,8,Obese,140/90,85,3000,Insomnia
7,Male,29,Teacher,6.3,6,40,7,Obese,140/90,82,3500,Insomnia
8,Male,29,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
9,Male,29,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
10,Male,29,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
11,Male,29,Doctor,6.1,6,30,8,Normal,120/80,70,8000,None
12,Male,29,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
13,Male,29,Doctor,6.1,6,30,8,Normal,120/80,70,8000,None
14,Male,29,Doctor,6,6,30,8,Normal,120/80,70,8000,None
15,Male,29,Doctor,6,6,30,8,Normal,120/80,70,8000,None
16,Male,29,Doctor,6,6,30,8,Normal,120/80,70,8000,None
17,Female,29,Nurse,6.5,5,40,7,Normal Weight,132/87,80,4000,Sleep Apnea
18,Male,29,Doctor,6,6,30,8,Normal,120/80,70,8000,Sleep Apnea
19,Female,29,Nurse,6.5,5,40,7,Normal Weight,132/87,80,4000,Insomnia
20,Male,30,Doctor,7.6,7,75,6,Normal,120/80,70,8000,None
21,Male,30,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
22,Male,30,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
23,Male,30,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
24,Male,30,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
25,Male,30,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
26,Male,30,Doctor,7.9,7,75,6,Normal,120/80,70,8000,None
27,Male,30,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
28,Male,30,Doctor,7.9,7,75,6,Normal,120/80,70,8000,None
29,Male,30,Doctor,7.9,7,75,6,Normal,120/80,70,8000,None
30,Male,30,Doctor,7.9,7,75,6,Normal,120/80,70,8000,None
31,Female,30,Nurse,6.4,5,35,7,Normal Weight,130/86,78,4100,Sleep Apnea
32,Female,30,Nurse,6.4,5,35,7,Normal Weight,130/86,78,4100,Insomnia
33,Female,31,Nurse,7.9,8,75,4,Normal Weight,117/76,69,6800,None
34,Male,31,Doctor,6.1,6,30,8,Normal,125/80,72,5000,None
35,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
36,Male,31,Doctor,6.1,6,30,8,Normal,125/80,72,5000,None
37,Male,31,Doctor,6.1,6,30,8,Normal,125/80,72,5000,None
38,Male,31,Doctor,7.6,7,75,6,Normal,120/80,70,8000,None
39,Male,31,Doctor,7.6,7,75,6,Normal,120/80,70,8000,None
40,Male,31,Doctor,7.6,7,75,6,Normal,120/80,70,8000,None
41,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
42,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
43,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
44,Male,31,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
45,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
46,Male,31,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
47,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
48,Male,31,Doctor,7.8,7,75,6,Normal,120/80,70,8000,None
49,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
50,Male,31,Doctor,7.7,7,75,6,Normal,120/80,70,8000,Sleep Apnea
51,Male,32,Engineer,7.5,8,45,3,Normal,120/80,70,8000,None
52,Male,32,Engineer,7.5,8,45,3,Normal,120/80,70,8000,None
53,Male,32,Doctor,6,6,30,8,Normal,125/80,72,5000,None
54,Male,32,Doctor,7.6,7,75,6,Normal,120/80,70,8000,None
55,Male,32,Doctor,6,6,30,8,Normal,125/80,72,5000,None
56,Male,32,Doctor,6,6,30,8,Normal,125/80,72,5000,None
57,Male,32,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
58,Male,32,Doctor,6,6,30,8,Normal,125/80,72,5000,None
59,Male,32,Doctor,6,6,30,8,Normal,125/80,72,5000,None
60,Male,32,Doctor,7.7,7,75,6,Normal,120/80,70,8000,None
61,Male,32,Doctor,6,6,30,8,Normal,125/80,72,5000,None
62,Male,32,Doctor,6,6,30,8,Normal,125/80,72,5000,None
63,Male,32,Doctor,6.2,6,30,8,Normal,125/80,72,5000,None
64,Male,32,Doctor,6.2,6,30,8,Normal,125/80,72,5000,None
65,Male,32,Doctor,6.2,6,30,8,Normal,125/80,72,5000,None
66,Male,32,Doctor,6.2,6,30,8,Normal,125/80,72,5000,None
67,Male,32,Accountant,7.2,8,50,6,Normal Weight,118/76,68,7000,None
68,Male,33,Doctor,6,6,30,8,Normal,125/80,72,5000,Insomnia
69,Female,33,Scientist,6.2,6,50,6,Overweight,128/85,76,5500,None
70,Female,33,Scientist,6.2,6,50,6,Overweight,128/85,76,5500,None
71,Male,33,Doctor,6.1,6,30,8,Normal,125/80,72,5000,None
72,Male,33,Doctor,6.1,6,30,8,Normal,125/80,72,5000,None
73,Male,33,Doctor,6.1,6,30,8,Normal,125/80,72,5000,None
74,Male,33,Doctor,6.1,6,30,8,Normal,125/80,72,5000,None
75,Male,33,Doctor,6,6,30,8,Normal,125/80,72,5000,None
76,Male,33,Doctor,6,6,30,8,Normal,125/80,72,5000,None
77,Male,33,Doctor,6,6,30,8,Normal,125/80,72,5000,None
78,Male,33,Doctor,6,6,30,8,Normal,125/80,72,5000,None
79,Male,33,Doctor,6,6,30,8,Normal,125/80,72,5000,None
80,Male,33,Doctor,6,6,30,8,Normal,125/80,72,5000,None
81,Female,34,Scientist,5.8,4,32,8,Overweight,131/86,81,5200,Sleep Apnea
82,Female,34,Scientist,5.8,4,32,8,Overweight,131/86,81,5200,Sleep Apnea
83,Male,35,Teacher,6.7,7,40,5,Overweight,128/84,70,5600,None
84,Male,35,Teacher,6.7,7,40,5,Overweight,128/84,70,5600,None
85,Male,35,Software Engineer,7.5,8,60,5,Normal Weight,120/80,70,8000,None
86,Female,35,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
87,Male,35,Engineer,7.2,8,60,4,Normal,125/80,65,5000,None
88,Male,35,Engineer,7.2,8,60,4,Normal,125/80,65,5000,None
89,Male,35,Engineer,7.3,8,60,4,Normal,125/80,65,5000,None
90,Male,35,Engineer,7.3,8,60,4,Normal,125/80,65,5000,None
91,Male,35,Engineer,7.3,8,60,4,Normal,125/80,65,5000,None
92,Male,35,Engineer,7.3,8,60,4,Normal,125/80,65,5000,None
93,Male,35,Software Engineer,7.5,8,60,5,Normal Weight,120/80,70,8000,None
94,Male,35,Lawyer,7.4,7,60,5,Obese,135/88,84,3300,Sleep Apnea
95,Female,36,Accountant,7.2,8,60,4,Normal,115/75,68,7000,Insomnia
96,Female,36,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
97,Female,36,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
98,Female,36,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
99,Female,36,Teacher,7.1,8,60,4,Normal,115/75,68,7000,None
100,Female,36,Teacher,7.1,8,60,4,Normal,115/75,68,7000,None
101,Female,36,Teacher,7.2,8,60,4,Normal,115/75,68,7000,None
102,Female,36,Teacher,7.2,8,60,4,Normal,115/75,68,7000,None
103,Female,36,Teacher,7.2,8,60,4,Normal,115/75,68,7000,None
104,Male,36,Teacher,6.6,5,35,7,Overweight,129/84,74,4800,Sleep Apnea
105,Female,36,Teacher,7.2,8,60,4,Normal,115/75,68,7000,Sleep Apnea
106,Male,36,Teacher,6.6,5,35,7,Overweight,129/84,74,4800,Insomnia
107,Female,37,Nurse,6.1,6,42,6,Overweight,126/83,77,4200,None
108,Male,37,Engineer,7.8,8,70,4,Normal Weight,120/80,68,7000,None
109,Male,37,Engineer,7.8,8,70,4,Normal Weight,120/80,68,7000,None
110,Male,37,Lawyer,7.4,8,60,5,Normal,130/85,68,8000,None
111,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
112,Male,37,Lawyer,7.4,8,60,5,Normal,130/85,68,8000,None
113,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
114,Male,37,Lawyer,7.4,8,60,5,Normal,130/85,68,8000,None
115,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
116,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
117,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
118,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
119,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
120,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
121,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
122,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
123,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
124,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
125,Female,37,Accountant,7.2,8,60,4,Normal,115/75,68,7000,None
126,Female,37,Nurse,7.5,8,60,4,Normal Weight,120/80,70,8000,None
127,Male,38,Lawyer,7.3,8,60,5,Normal,130/85,68,8000,None
128,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
129,Male,38,Lawyer,7.3,8,60,5,Normal,130/85,68,8000,None
130,Male,38,Lawyer,7.3,8,60,5,Normal,130/85,68,8000,None
131,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
132,Male,38,Lawyer,7.3,8,60,5,Normal,130/85,68,8000,None
133,Male,38,Lawyer,7.3,8,60,5,Normal,130/85,68,8000,None
134,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
135,Male,38,Lawyer,7.3,8,60,5,Normal,130/85,68,8000,None
136,Male,38,Lawyer,7.3,8,60,5,Normal,130/85,68,8000,None
137,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
138,Male,38,Lawyer,7.1,8,60,5,Normal,130/85,68,8000,None
139,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
140,Male,38,Lawyer,7.1,8,60,5,Normal,130/85,68,8000,None
141,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
142,Male,38,Lawyer,7.1,8,60,5,Normal,130/85,68,8000,None
143,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
144,Female,38,Accountant,7.1,8,60,4,Normal,115/75,68,7000,None
145,Male,38,Lawyer,7.1,8,60,5,Normal,130/85,68,8000,Sleep Apnea
146,Female,38,Lawyer,7.4,7,60,5,Obese,135/88,84,3300,Sleep Apnea
147,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,Insomnia
148,Male,39,Engineer,6.5,5,40,7,Overweight,132/87,80,4000,Insomnia
149,Female,39,Lawyer,6.9,7,50,6,Normal Weight,128/85,75,5500,None
150,Female,39,Accountant,8,9,80,3,Normal Weight,115/78,67,7500,None
151,Female,39,Accountant,8,9,80,3,Normal Weight,115/78,67,7500,None
152,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
153,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
154,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
155,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
156,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
157,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
158,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
159,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
160,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
161,Male,39,Lawyer,7.2,8,60,5,Normal,130/85,68,8000,None
162,Female,40,Accountant,7.2,8,55,6,Normal Weight,119/77,73,7300,None
163,Female,40,Accountant,7.2,8,55,6,Normal Weight,119/77,73,7300,None
164,Male,40,Lawyer,7.9,8,90,5,Normal,130/85,68,8000,None
165,Male,40,Lawyer,7.9,8,90,5,Normal,130/85,68,8000,None
166,Male,41,Lawyer,7.6,8,90,5,Normal,130/85,70,8000,Insomnia
167,Male,41,Engineer,7.3,8,70,6,Normal Weight,121/79,72,6200,None
168,Male,41,Lawyer,7.1,7,55,6,Overweight,125/82,72,6000,None
169,Male,41,Lawyer,7.1,7,55,6,Overweight,125/82,72,6000,None
170,Male,41,Lawyer,7.7,8,90,5,Normal,130/85,70,8000,None
171,Male,41,Lawyer,7.7,8,90,5,Normal,130/85,70,8000,None
172,Male,41,Lawyer,7.7,8,90,5,Normal,130/85,70,8000,None
173,Male,41,Lawyer,7.7,8,90,5,Normal,130/85,70,8000,None
174,Male,41,Lawyer,7.7,8,90,5,Normal,130/85,70,8000,None
175,Male,41,Lawyer,7.6,8,90,5,Normal,130/85,70,8000,None
176,Male,41,Lawyer,7.6,8,90,5,Normal,130/85,70,8000,None
177,Male,41,Lawyer,7.6,8,90,5,Normal,130/85,70,8000,None
178,Male,42,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
179,Male,42,Lawyer,7.8,8,90,5,Normal,130/85,70,8000,None
180,Male,42,Lawyer,7.8,8,90,5,Normal,130/85,70,8000,None
181,Male,42,Lawyer,7.8,8,90,5,Normal,130/85,70,8000,None
182,Male,42,Lawyer,7.8,8,90,5,Normal,130/85,70,8000,None
183,Male,42,Lawyer,7.8,8,90,5,Normal,130/85,70,8000,None
184,Male,42,Lawyer,7.8,8,90,5,Normal,130/85,70,8000,None
185,Female,42,Teacher,6.8,6,45,7,Overweight,130/85,78,5000,Sleep Apnea
186,Female,42,Teacher,6.8,6,45,7,Overweight,130/85,78,5000,Sleep Apnea
187,Female,43,Teacher,6.7,7,45,4,Overweight,135/90,65,6000,Insomnia
188,Male,43,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
189,Female,43,Teacher,6.7,7,45,4,Overweight,135/90,65,6000,Insomnia
190,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
191,Female,43,Teacher,6.7,7,45,4,Overweight,135/90,65,6000,Insomnia
192,Male,43,Salesperson,6.4,6,45,7,Overweight,130/85,72,6000,Insomnia
193,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
194,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
195,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
196,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
197,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
198,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
199,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
200,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
201,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Insomnia
202,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,Insomnia
203,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,Insomnia
204,Male,43,Engineer,6.9,6,47,7,Normal Weight,117/76,69,6800,None
205,Male,43,Engineer,7.6,8,75,4,Overweight,122/80,68,6800,None
206,Male,43,Engineer,7.7,8,90,5,Normal,130/85,70,8000,None
207,Male,43,Engineer,7.7,8,90,5,Normal,130/85,70,8000,None
208,Male,43,Engineer,7.7,8,90,5,Normal,130/85,70,8000,None
209,Male,43,Engineer,7.7,8,90,5,Normal,130/85,70,8000,None
210,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
211,Male,43,Engineer,7.7,8,90,5,Normal,130/85,70,8000,None
212,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
213,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
214,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
215,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
216,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
217,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
218,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,None
219,Male,43,Engineer,7.8,8,90,5,Normal,130/85,70,8000,Sleep Apnea
220,Male,43,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,Sleep Apnea
221,Female,44,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
222,Male,44,Salesperson,6.4,6,45,7,Overweight,130/85,72,6000,Insomnia
223,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
224,Male,44,Salesperson,6.4,6,45,7,Overweight,130/85,72,6000,Insomnia
225,Female,44,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
226,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
227,Female,44,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
228,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
229,Female,44,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
230,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
231,Female,44,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
232,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
233,Female,44,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
234,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
235,Female,44,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
236,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
237,Male,44,Salesperson,6.4,6,45,7,Overweight,130/85,72,6000,Insomnia
238,Female,44,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
239,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
240,Male,44,Salesperson,6.4,6,45,7,Overweight,130/85,72,6000,Insomnia
241,Female,44,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
242,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
243,Male,44,Salesperson,6.4,6,45,7,Overweight,130/85,72,6000,Insomnia
244,Female,44,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
245,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
246,Female,44,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
247,Male,44,Salesperson,6.3,6,45,7,Overweight,130/85,72,6000,Insomnia
248,Male,44,Engineer,6.8,7,45,7,Overweight,130/85,78,5000,Insomnia
249,Male,44,Salesperson,6.4,6,45,7,Overweight,130/85,72,6000,None
250,Male,44,Salesperson,6.5,6,45,7,Overweight,130/85,72,6000,None
251,Female,45,Teacher,6.8,7,30,6,Overweight,135/90,65,6000,Insomnia
252,Female,45,Teacher,6.8,7,30,6,Overweight,135/90,65,6000,Insomnia
253,Female,45,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
254,Female,45,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
255,Female,45,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
256,Female,45,Teacher,6.5,7,45,4,Overweight,135/90,65,6000,Insomnia
257,Female,45,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
258,Female,45,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
259,Female,45,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
260,Female,45,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
261,Female,45,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,Insomnia
262,Female,45,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,None
263,Female,45,Teacher,6.6,7,45,4,Overweight,135/90,65,6000,None
264,Female,45,Manager,6.9,7,55,5,Overweight,125/82,75,5500,None
265,Male,48,Doctor,7.3,7,65,5,Obese,142/92,83,3500,Insomnia
266,Female,48,Nurse,5.9,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
267,Male,48,Doctor,7.3,7,65,5,Obese,142/92,83,3500,Insomnia
268,Female,49,Nurse,6.2,6,90,8,Overweight,140/95,75,10000,None
269,Female,49,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
270,Female,49,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
271,Female,49,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
272,Female,49,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
273,Female,49,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
274,Female,49,Nurse,6.2,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
275,Female,49,Nurse,6.2,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
276,Female,49,Nurse,6.2,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
277,Male,49,Doctor,8.1,9,85,3,Obese,139/91,86,3700,Sleep Apnea
278,Male,49,Doctor,8.1,9,85,3,Obese,139/91,86,3700,Sleep Apnea
279,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Insomnia
280,Female,50,Engineer,8.3,9,30,3,Normal,125/80,65,5000,None
281,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,None
282,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
283,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
284,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
285,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
286,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
287,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
288,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
289,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
290,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
291,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
292,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
293,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
294,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
295,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
296,Female,50,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
297,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
298,Female,50,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
299,Female,51,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
300,Female,51,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
301,Female,51,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
302,Female,51,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
303,Female,51,Nurse,7.1,7,55,6,Normal Weight,125/82,72,6000,None
304,Female,51,Nurse,6,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
305,Female,51,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
306,Female,51,Nurse,6.1,6,90,8,Overweight,140/95,75,10000,Sleep Apnea
307,Female,52,Accountant,6.5,7,45,7,Overweight,130/85,72,6000,Insomnia
308,Female,52,Accountant,6.5,7,45,7,Overweight,130/85,72,6000,Insomnia
309,Female,52,Accountant,6.6,7,45,7,Overweight,130/85,72,6000,Insomnia
310,Female,52,Accountant,6.6,7,45,7,Overweight,130/85,72,6000,Insomnia
311,Female,52,Accountant,6.6,7,45,7,Overweight,130/85,72,6000,Insomnia
312,Female,52,Accountant,6.6,7,45,7,Overweight,130/85,72,6000,Insomnia
313,Female,52,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
314,Female,52,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
315,Female,52,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
316,Female,53,Engineer,8.3,9,30,3,Normal,125/80,65,5000,Insomnia
317,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
318,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
319,Female,53,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
320,Female,53,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
321,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
322,Female,53,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
323,Female,53,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
324,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
325,Female,53,Engineer,8.3,9,30,3,Normal,125/80,65,5000,None
326,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
327,Female,53,Engineer,8.3,9,30,3,Normal,125/80,65,5000,None
328,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
329,Female,53,Engineer,8.3,9,30,3,Normal,125/80,65,5000,None
330,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
331,Female,53,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
332,Female,53,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
333,Female,54,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
334,Female,54,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
335,Female,54,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
336,Female,54,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
337,Female,54,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
338,Female,54,Engineer,8.4,9,30,3,Normal,125/80,65,5000,None
339,Female,54,Engineer,8.5,9,30,3,Normal,125/80,65,5000,None
340,Female,55,Nurse,8.1,9,75,4,Overweight,140/95,72,5000,Sleep Apnea
341,Female,55,Nurse,8.1,9,75,4,Overweight,140/95,72,5000,Sleep Apnea
342,Female,56,Doctor,8.2,9,90,3,Normal Weight,118/75,65,10000,None
343,Female,56,Doctor,8.2,9,90,3,Normal Weight,118/75,65,10000,None
344,Female,57,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,None
345,Female,57,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
346,Female,57,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
347,Female,57,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
348,Female,57,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
349,Female,57,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
350,Female,57,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
351,Female,57,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
352,Female,57,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
353,Female,58,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
354,Female,58,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
355,Female,58,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
356,Female,58,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
357,Female,58,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
358,Female,58,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
359,Female,59,Nurse,8,9,75,3,Overweight,140/95,68,7000,None
360,Female,59,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,None
361,Female,59,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
362,Female,59,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
363,Female,59,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
364,Female,59,Nurse,8.2,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
365,Female,59,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
366,Female,59,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
367,Female,59,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
368,Female,59,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
369,Female,59,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
370,Female,59,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
371,Female,59,Nurse,8,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
372,Female,59,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
373,Female,59,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea
374,Female,59,Nurse,8.1,9,75,3,Overweight,140/95,68,7000,Sleep Apnea

OpenAI Chat Completion v1

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY")
)

prompt = "who is best ronaldo or messi?"

def chat_comp(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt},
                  {"role": "system", "content": "You are a super hero!"}
                  ],
        max_tokens=250
    )

    return response.choices[0].message.content

print(chat_comp(prompt))

Generate a joke using ChatGPT

from bentoml import client
from openai import OpenAI
import os

#client = OpenAI(
#    api_key=os.environ.get("OPENAI_API_KEY")
#)

client  = OpenAI()

prompt = "tell me a joke!"

def chat_comp(prompt):

    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=250,
        temperature=0.7
    )

    return response.choices[0].message.content

print(chat_comp(prompt))

Chat completions shell script

(.venv) bharathkumardasaraju@demos$ python3.12 ./chat_completion1.py
Making more money often involves exploring various strategies based on your skills, resources, and interests. Here are several approaches you can consider:

### 1. **Enhance Your Skills**
   - **Invest in Education:** Take courses or certifications in high-demand fields that offer higher pay.
   - **Learn New Skills:** Acquire skills related to technology, marketing, or finance, which can lead to better job opportunities.

### 2. **Negotiate Your Salary**
   - **Research Market Rates:** Know your worth in the job market and be prepared to negotiate your salary or ask for a raise.
   - **Prepare Your Case:** Highlight your contributions, achievements, and any additional responsibilities you’ve taken on.

### 3. **Explore Side Gigs**
   - **Freelancing:** Offer your skills on platforms like Upwork, Fiverr, or Freelancer.
   - **Part-Time Jobs:** Consider flexible options such as tutoring, ride-sharing, or food delivery.

### 4. **Invest Wisely**
   - **Stock Market:** Learn about investing in stocks, ETFs, or mutual funds. Understand the risks involved.
   - **Real Estate:** Explore rental properties or real estate investment trusts (REITs).

### 5. **Start
(.venv) bharathkumardasaraju@demos$ python3.12 ./chat_completion1.py
Making more money can involve a variety of strategies depending on your skills, resources, and circumstances. Here are some ideas to consider:

### Side Hustles
1. **Freelancing**: Use your skills (writing, graphic design, programming, etc.) on platforms like Upwork or Fiverr.
2. **Tutoring**: Offer your expertise in subjects you excel in, either online or in-person.
3. **Online Selling**: Sell products on Etsy, eBay, or Amazon. You can create handmade goods or curations of items you find.

### Investments
1. **Stock Market**: Invest in stocks or ETFs. Make sure to educate yourself about investing before diving in.
2. **Real Estate**: Consider rental properties or real estate crowdfunding platforms if you have some capital to start with.
3. **Cryptocurrency**: This can be highly volatile, but if you are knowledgeable, it can yield high returns.

### Passive Income
1. **Create an Online Course**: Share your knowledge through platforms like Udemy or Teachable.
2. **Write an E-book**: If you have expertise in a particular area, consider writing and self-publishing a book.
3. **Affiliate Marketing**: Promote products and
(.venv) bharathkumardasaraju@demos$





(.venv) bharathkumardasaraju@demos$ python3.12 ./tell_a_joke.py
Why don't skeletons fight each other?

They don't have the guts!
(.venv) bharathkumardasaraju@demos$

Updated joke generation script

from openai import OpenAI
import os
client = OpenAI(
    base_url=os.environ.get("OPENAI_API_BASE")
)
prompt="Tell me a joke"
response = client.chat.completions.create(
            model= "gpt-4o-mini",
            messages=[{"role":"user","content":prompt}],
            max_tokens=100,
            temperature=0.7)
print(response.choices[0].message.content)

OpenAI v1 completions shell script

~/openaiproject ✖ env | grep -i "OPENAI_"
OPENAI_API_KEY=Sk-kkAI-5b7c86e365f2794c6412ef3aecc82c6e6576aa756950e4b31389a6a9a46e8062kk_ogj3yy6h2rudgznp-kk6c284664
AZURE_OPENAI_API_KEY=Sk-kkAI-5b7c86e365f2794c6412ef3aecc82c6e6576aa756950e4b31389a6a9a46e8062kk_ogj3yy6h2rudgznp-kk6c284664
OPENAI_API_BASE=https://main.bk-ai-keys.bharathkumar.com/v1
AZURE_OPENAI_ENDPOINT=https://main.bk-ai-keys.bharathkumar.com/openai

~/openaiproject ➜


 ~/openaiproject ➜  python3 tell_a_joke_new.py
Why did the scarecrow win an award?

Because he was outstanding in his field!

~/openaiproject ➜

Generate cooking recipes

from openai import OpenAI

client  = OpenAI()

# Create something that can accept ingredients
ingredients = []

while True:
    ingredient = input("Enter ingredients. Type done once complete:")
    if ingredient.lower() == "done":
        break

    ingredients.append(ingredient)

# Create a function that uses the model and brings in the ingredients

def recipe_gen(ingredients):
    messages = []
    for ingredient in ingredients:
        messages.append({"role": "user", "content": ingredient})

    messages.extend([
        {"role": "system", "content": "direct, point"},
        {"role": "assistant", "content": "You are a high-end chef, Generate a recipe based on the given ingredients."}]
    )
    response = client.chat.completions.create(
        model="gpt-4o",
        messages= messages,
        max_tokens=300,
        temperature=0.9
    )

    return response.choices[0].message.content


print(recipe_gen(ingredients))

Script to generate food recipes

(.venv) bharathkumardasaraju@demos$ python3.12 ./recipe_generator.py
Enter ingredients. Type done once complete:apples
Enter ingredients. Type done once complete:sugar
Enter ingredients. Type done once complete:grapes
Enter ingredients. Type done once complete:chicken
Enter ingredients. Type done once complete:done
Traceback (most recent call last):
  File "/Users/bharathkumardasaraju/external/learn-ml-ops/11.OpenAI/3.TextGeneration/demos/./recipe_generator.py", line 36, in <module>
    print(recipe_gen(ingredients))
          ^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/bharathkumardasaraju/external/learn-ml-ops/11.OpenAI/3.TextGeneration/demos/./recipe_generator.py", line 22, in recipe_gen
    messages.create(
    ^^^^^^^^^^^^^^^
AttributeError: 'list' object has no attribute 'create'
(.venv) bharathkumardasaraju@demos$




(.venv) bharathkumardasaraju@demos$ python3.12 ./recipe_generator.py
Enter ingredients. Type done once complete:grapes
Enter ingredients. Type done once complete:milk
Enter ingredients. Type done once complete:beef
Enter ingredients. Type done once complete:chocolate
Enter ingredients. Type done once complete:sugar
Enter ingredients. Type done once complete:done
Here is a delicious recipe that incorporates grapes, milk, beef, chocolate, and sugar:

**Chocolate-Glazed Beef with Grape Reduction**

**Ingredients:**
- 1 lb beef tenderloin or steak
- Salt and pepper to taste
- 2 tbsp olive oil
- 1 cup red seedless grapes, halved
- 1/2 cup red wine
- 1 cup beef stock
- 1/4 cup milk
- 2 oz dark chocolate, chopped
- 2 tbsp sugar
- Fresh thyme or rosemary for garnish (optional)

**Instructions:**

1. **Prepare the Beef:**
   - Season the beef tenderloin with salt and pepper.
   - In a large skillet, heat olive oil over medium-high heat. Sear the beef on all sides until browned, about 3-4 minutes per side. Transfer the beef to a plate and cover with foil to keep warm.

2. **Make the Grape Reduction:**
   - In the same skillet, add the grape halves and cook for 2-3 minutes until they begin to soften.
   - Pour in the red wine and let it simmer for about 5 minutes until the liquid reduces by half.
   - Add the beef stock and continue to simmer for another 5 minutes. The reduction should thicken slightly.

3. **Create the Chocolate Glaze:**
   - In a small saucepan, heat the milk over low heat.
   - Add the chopped dark
(.venv) bharathkumardasaraju@demos$

Translate content with OpenAI

from openai import OpenAI

client = OpenAI()

# article
article = "ప్రతిపత్తిస్వత్వముల విషయమున మానవులెల్లరును జన్మతః స్వతంత్రులును సమానులును నగుదురు. వారు వివేచన-అంతఃకరణ సంపన్నులగుటచే పరస్పరము భ్రాతృభావముతో వర్తింపవలయును."

# prompt
prompt = f"translate the following article: {article}"

# create a function that can translates the article
def article_translate(prompt):
    response = client.chat.completions.create(
        model = "gpt-4o",
        messages = [{"role": "user", "content": prompt},
                    {"role": "assistant", "content": "You are a professional translator. You translate news articles into English"},
                    {"role": "system", "content": "Direct english translator"}
                    ],
        temperature=0.1
    )
    return response.choices[0].message.content

print(article_translate(prompt))

Translation via OpenAI shell

Translating Article

(.venv) bharathkumardasaraju@demos$ python3.12 ./translate_article.py
All human beings are born free and equal in dignity and rights. They are endowed with reason and conscience and should act towards one another in a spirit of brotherhood.
(.venv) bharathkumardasaraju@demos$

Shell story generator with LLM

python3 -m venv /opt/venv
source /opt/venv/bin/activate
pip install openai


~ ➜  env | grep -i "OPENAI_"
OPENAI_API_KEY=Sk-kkAI-8f0922fb7d5c3d98caf6386f9c9c0bbacf4608280c1a5788d3ac7b52183b5ffckk_qo7zhbykltyd2dj6-kkcf74f528
AZURE_OPENAI_API_KEY=Sk-kkAI-8f0922fb7d5c3d98caf6386f9c9c0bbacf4608280c1a5788d3ac7b52183b5ffckk_qo7zhbykltyd2dj6-kkcf74f528
OPENAI_API_BASE=https://main.bk-ai-keys.bharathkumar.com/v1
AZURE_OPENAI_ENDPOINT=https://main.bk-ai-keys.bharathkumar.com/openai

~ ➜

~/openaiproject ➜  python3 short-story-gen.py
Prompt: Write a short story about a young knight who finds a mysterious map.
Generated Story:
In the quaint village of Eldergrove, nestled between emerald hills and whispering streams, there lived a young knight named Cedric. He was not yet of the stature of the famed knights of legend but was known for his bravery and kind heart. Every evening, after his training at the castle, Cedric would wander the dense forest that bordered Eldergrove, dreaming of great adventures and glory.

One crisp autumn afternoon, as golden leaves danced in the wind, Cedric stumbled upon a clearing he

Prompt: Tell a short story about an astronaut stranded on a distant planet.
Generated Story:
Captain Elara Voss had always dreamed of exploring the stars, her heart set on the uncharted expanses of the universe. When she was chosen for the mission to the distant planet of Theros-7, she felt as if she had finally reached the pinnacle of her aspirations. But a freak meteor shower during her approach had sent her spacecraft spiraling, forcing her to make an emergency landing on the planets surface.

As she emerged from the wreckage, the sun hung low on the horizon

Prompt: Generate a short story about a talking cat who solves mysteries.
Generated Story:
Once upon a time in the quaint little town of Whiskerville, there lived a peculiar cat named Sherlock. Unlike any other feline, Sherlock possessed the extraordinary ability to speak. With sleek, ebony fur and piercing green eyes, he was both charming and astute, often found lounging on the windowsill of his owner, an elderly woman named Mrs. Penelope.

Mrs. Penelope adored Sherlock, often saying, “You’re the smartest creature in Whiskerville!” Little did she know just how


~/openaiproject ➜

Short story generator

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),
    base_url=os.environ.get("OPENAI_API_BASE")
)

def generate_short_story(prompt):
    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100,
        temperature=0.7
    )
    return response.choices[0].message.content

prompts = [
                "Write a short story about a young knight who finds a mysterious map.",
                "Tell a short story about an astronaut stranded on a distant planet.",
                "Generate a short story about a talking cat who solves mysteries."]

for prompt in prompts:
    print(f"Prompt: {prompt}")
    story = generate_short_story(prompt)
    print(f"Generated Story:\n{story}\n")

User dataset for personalized training

identifierHash,type,country,language,socialNbFollowers,socialNbFollows,socialProductsLiked,productsListed,productsSold,productsPassRate,productsWished,productsBought,gender,civilityGenderId,civilityTitle,hasAnyApp,hasAndroidApp,hasIosApp,hasProfilePicture,daysSinceLastLogin,seniority,seniorityAsMonths,seniorityAsYears,countryCode
-1097895247965112460,user,Royaume-Uni,en,147,10,77,26,174,74.0,104,1,M,1,mr,True,False,True,True,11,3196,106.53,8.88,gb
2347567364561867620,user,Monaco,en,167,8,2,19,170,99.0,0,0,F,2,mrs,True,False,True,True,12,3204,106.8,8.9,mc
6870940546848049750,user,France,fr,137,13,60,33,163,94.0,10,3,F,2,mrs,True,False,True,False,11,3203,106.77,8.9,fr
-4640272621319568052,user,Etats-Unis,en,131,10,14,122,152,92.0,7,0,F,2,mrs,True,False,True,False,12,3198,106.6,8.88,us
-5175830994878542658,user,Etats-Unis,en,167,8,0,25,125,100.0,0,0,F,2,mrs,False,False,False,True,22,2854,95.13,7.93,us
7631788075812383072,user,Allemagne,de,130,12,1,47,123,91.0,0,0,F,2,mrs,True,False,True,False,11,3196,106.53,8.88,de
674361423306028463,user,Suède,en,121,0,1140,31,108,94.0,531,105,F,3,miss,True,True,False,False,11,3198,106.6,8.88,se
2550976450216757005,user,France,fr,53,9,3,5,106,98.0,0,0,F,2,mrs,True,False,True,True,11,2857,95.23,7.94,fr
3718185418791028367,user,Italie,it,744,13764,51671,0,104,85.0,1842,0,F,2,mrs,True,False,True,False,14,3195,106.5,8.88,it
3908244093584862523,user,Royaume-Uni,en,57,8,45,123,92,74.0,6,2,F,3,miss,True,False,True,True,11,2856,95.2,7.93,gb
-8291872437813427796,user,Italie,it,127,40,863,40,87,89.0,68,0,M,1,mr,True,False,True,False,11,3195,106.5,8.88,it
-7992812913368049944,user,Etats-Unis,en,102,8,0,0,87,98.0,0,0,F,2,mrs,False,False,False,True,68,2855,95.17,7.93,us
-6042837405271512617,user,Espagne,es,46,19,1308,63,87,92.0,564,36,F,2,mrs,True,False,True,True,11,3195,106.5,8.88,es
8027803513424684410,user,Italie,it,35,9,2,185,87,94.0,2,0,M,1,mr,True,False,True,False,11,2859,95.3,7.94,it
-1009317821982259467,user,Etats-Unis,en,123,8,1,117,86,100.0,0,0,F,2,mrs,False,False,False,True,55,2853,95.1,7.92,us
1205881756843387030,user,Italie,it,45,8,0,17,84,98.7,0,0,M,1,mr,True,False,True,True,11,3195,106.5,8.88,it
1450851698857145434,user,Allemagne,de,48,16,599,22,82,89.0,1016,32,F,2,mrs,True,False,True,False,11,3203,106.77,8.9,de
-9221801386431225948,user,Espagne,es,176,60,276,18,81,92.0,1,0,F,2,mrs,True,False,True,False,11,2858,95.27,7.94,es
732189832737494520,user,Royaume-Uni,en,47,8,0,10,78,98.0,0,0,F,2,mrs,True,False,True,True,13,3205,106.83,8.9,gb
8741445689531731651,user,Espagne,es,85,0,115,54,73,90.0,4,0,M,1,mr,True,True,False,False,11,3202,106.73,8.89,es
8317887111506061089,user,France,fr,107,3649,310,217,70,98.0,0,0,M,1,mr,True,False,True,True,11,3204,106.8,8.9,fr
5145876400632363206,user,France,fr,60,9,41,244,70,88.0,37,14,F,2,mrs,False,False,False,True,11,3197,106.57,8.88,fr
6048206785556031922,user,France,fr,73,27,104,8,69,98.0,182,115,F,2,mrs,True,False,True,False,11,2853,95.1,7.92,fr
3275083886092943154,user,Suède,en,125,500,7044,117,67,90.0,664,0,F,2,mrs,False,False,False,False,16,2855,95.17,7.93,se
5724841529455566712,user,Italie,it,103,431,7,13,66,93.0,0,0,F,2,mrs,True,False,True,False,15,3202,106.73,8.89,it
-7107267944201976220,user,France,fr,56,8,727,4,66,83.0,16,0,F,2,mrs,True,False,True,False,11,3195,106.5,8.88,fr
-7659168937010608427,user,Espagne,es,51,9,66,13,65,90.0,1,0,F,2,mrs,True,False,True,True,11,3195,106.5,8.88,es
-6663008189229430140,user,Croatie,en,50,8,0,44,64,93.0,0,0,F,2,mrs,False,False,False,True,11,2856,95.2,7.93,hr
-3666404711126006655,user,Etats-Unis,en,54,8,5,102,63,98.0,3,0,M,1,mr,False,False,False,True,293,3204,106.8,8.9,us
7927686682563456169,user,Italie,it,47,8,4,9,63,98.0,1,0,F,2,mrs,True,True,False,True,12,2857,95.23,7.94,it
5993803159989834567,user,France,fr,36,12,35,60,62,87.0,2,2,F,2,mrs,True,False,True,False,12,3196,106.53,8.88,fr
-6114870297213503871,user,Lettonie,en,40,189,68,15,60,90.0,2,0,F,2,mrs,False,False,False,False,11,3200,106.67,8.89,lv
-5736812074957181748,user,Royaume-Uni,en,23,22,28,23,60,96.4,139,6,F,2,mrs,False,False,False,False,11,2855,95.17,7.93,gb
-1729009415811639067,user,Taiwan,en,83,8,4,56,57,98.0,0,0,F,2,mrs,True,False,True,True,11,2855,95.17,7.93,tw
-697661860180574372,user,Royaume-Uni,en,52,8,0,32,55,100.0,0,0,F,2,mrs,False,False,False,True,86,3205,106.83,8.9,gb
2555419810142554931,user,Royaume-Uni,en,42,8,0,55,54,100.0,0,0,F,2,mrs,False,False,False,True,23,3203,106.77,8.9,gb
-6017183581142815690,user,Royaume-Uni,en,28,8,83,34,51,100.0,25,8,F,2,mrs,True,False,True,True,12,2854,95.13,7.93,gb
-6184929557711343088,user,Espagne,es,96,0,3,0,50,96.0,214,32,F,2,mrs,True,False,True,False,13,3201,106.7,8.89,es
751170060116305027,user,Italie,it,58,8,2,62,50,100.0,0,0,M,1,mr,False,False,False,False,11,3195,106.5,8.88,it
866533151187667584,user,Espagne,es,53,9,4,38,50,68.0,3,0,F,2,mrs,True,False,True,True,240,2857,95.23,7.94,es
2063789019158959186,user,Italie,it,63,8,2,7,49,86.4,0,0,F,2,mrs,True,True,False,False,12,3202,106.73,8.89,it
-6200949683536787250,user,Bulgarie,en,30,8,7,6,48,35.0,0,0,F,2,mrs,True,False,True,False,43,3204,106.8,8.9,bg
8848554873545217386,user,Espagne,es,58,48,593,39,47,93.0,66,1,F,2,mrs,True,False,True,False,11,2854,95.13,7.93,es
3479861271803167547,user,Royaume-Uni,en,28,8,4,0,47,77.0,8,1,F,2,mrs,True,False,True,False,21,3198,106.6,8.88,gb
2624921703073218682,user,Etats-Unis,en,53,61,1160,62,45,80.0,4,69,F,2,mrs,True,False,True,False,11,2855,95.17,7.93,us
6316993533074981913,user,France,fr,22,8,0,2,44,89.1,0,0,F,2,mrs,True,True,False,True,11,3200,106.67,8.89,fr
4006961445353148761,user,Italie,it,21,8,12,7,44,93.0,9,3,M,1,mr,True,False,True,False,11,3196,106.53,8.88,it
-1006797016053203392,user,Italie,it,47,8,13,25,43,93.0,19,0,F,2,mrs,True,False,True,False,11,3195,106.5,8.88,it
1243326618075091756,user,France,en,41,8,0,23,43,78.0,0,0,F,2,mrs,True,True,True,False,11,3203,106.77,8.9,fr
-1476665525462223230,user,Bulgarie,en,67,30,345,4,42,79.0,157,12,F,2,mrs,True,False,True,False,11,3199,106.63,8.89,bg
-685666612473067274,user,Italie,it,100,450,544,82,41,90.0,114,9,F,2,mrs,True,False,True,False,11,2856,95.2,7.93,it
-3509281294387690066,user,Chypre,en,39,17,474,66,41,87.0,17,0,F,2,mrs,False,False,False,False,11,2852,95.07,7.92,cy
-3240410723839501801,user,Royaume-Uni,en,9,8,3,5,41,78.0,1,1,M,1,mr,True,False,True,True,11,2859,95.3,7.94,gb
-7003214344574182759,user,France,fr,15,8,0,10,38,84.0,1,0,F,2,mrs,False,False,False,True,11,3198,106.6,8.88,fr
245123921236534929,user,Royaume-Uni,en,82,12,150,5,37,50.0,7,5,F,3,miss,True,False,True,False,106,2857,95.23,7.94,gb
3603032965600555065,user,Italie,it,40,8,0,12,37,94.0,0,0,F,2,mrs,True,False,True,False,11,2856,95.2,7.93,it
5888568966094757110,user,France,fr,111,37,237,7,35,91.0,2,1,F,2,mrs,True,False,True,False,11,2853,95.1,7.92,fr
-674375683452639256,user,Italie,it,172,50,778,96,34,64.0,2,0,F,2,mrs,True,False,True,False,11,3202,106.73,8.89,it
-1518127094662515469,user,Italie,it,43,12,204,40,34,96.2,1,0,F,2,mrs,True,True,True,False,12,2858,95.27,7.94,it
414557244265065039,user,Italie,it,24,23,34,46,34,85.0,9,1,F,2,mrs,True,False,True,False,11,3197,106.57,8.88,it
-2593705224835681064,user,Italie,en,94,9,5,10,33,28.0,0,0,M,1,mr,True,True,True,False,77,2858,95.27,7.94,it
-3756574628479963692,user,Allemagne,de,30,8,16,13,33,96.0,97,1,F,2,mrs,False,False,False,False,12,2857,95.23,7.94,de
-4708792591984924121,user,France,fr,18,9,4,7,33,93.0,2,0,F,2,mrs,True,False,True,True,16,3196,106.53,8.88,fr
-2423762989463308126,user,Italie,it,16,17,14,24,33,87.0,0,0,F,2,mrs,True,False,True,True,13,3196,106.53,8.88,it
4878417912154087005,user,Bulgarie,en,57,11,26,55,31,100.0,3,0,F,2,mrs,True,False,True,True,11,3204,106.8,8.9,bg
-3105560726122001062,user,Espagne,es,17,9,16,8,31,90.0,1,0,M,1,mr,True,False,True,True,17,3200,106.67,8.89,es
2471908634811746774,user,France,fr,17,8,1,35,31,93.0,2,0,F,2,mrs,False,False,False,True,11,2858,95.27,7.94,fr
-707350252975782312,user,Luxembourg,en,52,3,137,0,30,90.0,117,80,F,2,mrs,True,False,True,False,11,3199,106.63,8.89,lu
-266967029948851592,user,France,fr,48,8,9,0,30,93.0,1,0,F,2,mrs,True,True,False,False,131,2858,95.27,7.94,fr
-5126039730159284591,user,Royaume-Uni,en,31,8,84,4,30,100.0,19,0,F,2,mrs,True,False,True,True,14,3200,106.67,8.89,gb
5040679896375816853,user,Italie,it,24,8,27,113,30,76.0,9,1,F,2,mrs,True,False,True,False,11,2857,95.23,7.94,it
7628056793489694733,user,Italie,it,20,8,11,12,30,93.0,0,1,F,2,mrs,True,False,True,True,12,3198,106.6,8.88,it
1148664357418862256,user,Italie,it,19,8,1,15,30,83.0,0,1,F,2,mrs,True,False,True,True,13,3203,106.77,8.9,it
-7525201869937988415,user,Royaume-Uni,en,7,8,0,37,30,90.0,0,0,F,2,mrs,True,False,True,False,13,2853,95.1,7.92,gb
-7305545168149330332,user,Royaume-Uni,en,42,8,0,5,29,86.0,0,0,F,2,mrs,True,False,True,True,15,3197,106.57,8.88,gb
-4989163984886404589,user,Royaume-Uni,en,21,8,1,5,29,100.0,0,0,F,2,mrs,True,False,True,True,198,3205,106.83,8.9,gb
-5487622799428918401,user,Italie,it,18,9,19,84,29,66.0,15,0,F,2,mrs,True,False,True,True,11,3197,106.57,8.88,it
5578087612076626342,user,Italie,it,38,9,0,5,28,96.0,0,0,M,1,mr,True,False,True,False,14,3201,106.7,8.89,it
460632335991029855,user,France,fr,28,8,0,0,28,89.0,0,0,F,2,mrs,True,False,True,True,15,3200,106.67,8.89,fr
-3031036438332403639,user,Espagne,es,25,1,0,32,28,89.0,0,0,F,2,mrs,True,False,True,False,12,2853,95.1,7.92,es
-3225547952260578756,user,Italie,it,18,8,16,4,28,100.0,0,0,F,2,mrs,True,False,True,False,11,3195,106.5,8.88,it
-2019638691506243097,user,Roumanie,en,13,8,3,16,28,96.0,0,0,F,2,mrs,False,False,False,True,11,2853,95.1,7.92,ro
2447725262541190866,user,Suède,en,62,16,91,8,27,41.6,106,8,M,1,mr,True,False,True,False,11,2859,95.3,7.94,se
-475961064526915713,user,Italie,it,37,9,122,23,27,85.0,0,0,M,1,mr,True,False,True,False,11,3202,106.73,8.89,it
-5265011381291166158,user,Pays-Bas,en,35,11,8,2,27,55.0,1,0,M,1,mr,True,False,True,False,30,3202,106.73,8.89,nl
392351478755041771,user,France,fr,34,9,46,25,27,96.0,0,0,F,2,mrs,True,False,True,False,12,2855,95.17,7.93,fr
-155472512863609170,user,Italie,it,21,9,3,20,27,85.0,0,0,F,3,miss,True,True,True,False,11,3204,106.8,8.9,it
-7333104076460973307,user,Slovaquie,en,15,8,9,14,27,74.0,0,0,M,1,mr,True,False,True,False,15,3197,106.57,8.88,sk
6362794750925083140,user,Royaume-Uni,en,14,8,0,3,27,100.0,0,0,F,2,mrs,False,False,False,True,62,3205,106.83,8.9,gb
2296455035567004619,user,Royaume-Uni,en,29,8,2,13,26,100.0,0,0,M,1,mr,False,False,False,True,11,2853,95.1,7.92,gb
-1558789575290938745,user,Italie,it,28,9,11,17,26,82.6,0,0,M,1,mr,True,False,True,False,11,3195,106.5,8.88,it
-118824317870262106,user,Italie,it,25,10,0,1,26,100.0,0,0,F,2,mrs,False,False,False,True,40,3195,106.5,8.88,it
5570386793953288474,user,Espagne,es,15,8,1,17,26,88.0,0,0,F,2,mrs,True,True,False,True,20,3197,106.57,8.88,es
4039848352999540316,user,Italie,it,10,8,5,48,26,95.0,3,1,F,2,mrs,True,False,True,True,11,3199,106.63,8.89,it
6198034727570263587,user,Royaume-Uni,en,52,8,760,18,24,37.0,182,13,F,2,mrs,True,False,True,False,12,3199,106.63,8.89,gb
-3767129840327519972,user,Italie,it,34,16,35,28,24,84.0,1,0,F,2,mrs,True,True,False,False,11,3204,106.8,8.9,it
-3642418212370400810,user,Bulgarie,en,31,27,161,4,24,100.0,612,20,F,2,mrs,True,False,True,True,11,3198,106.6,8.88,bg
-4289690996070696899,user,Espagne,es,31,8,6,1,24,76.0,0,0,F,2,mrs,False,False,False,False,13,2858,95.27,7.94,es
-2032563623176529528,user,Italie,it,27,8,0,1,24,90.9,0,0,F,2,mrs,False,False,False,True,38,2856,95.2,7.93,it

OpenAI for AI research assistant

import os
import pandas as pd
from openai import OpenAI

# Read API key from environment
api_key = os.getenv("OPENAI_API_KEY")

# Initialize OpenAI client
client = OpenAI(api_key=api_key)

# Read the CSV file
csv_file_path = "./users.csv"
df = pd.read_csv(csv_file_path)

# Define a function to analyze data using OpenAI
def analyze_data(df):
    response = client.chat.completions.create(
        model="gpt-3.5-turbo",
        messages=[
            {
                "role": "user",
                "content": f"You are a research assistant. Provide key insights from the following dataset in point form only from demographics :\n\n{df.to_string()}"
            }
        ],
        max_tokens=500,
        temperature=0.2
    )

    return response.choices[0].message.content

# Run analysis and print insights
print(analyze_data(df))

Research assistant shell script

(.venv) bharathkumardasaraju@demos$ python3.12 ./ai_research_assistant.py
- The dataset contains information about users from different countries.
- Most users have social media followers and follow others.
- The majority of users have listed products for sale.
- There is variation in the number of products sold by users.
- The pass rate for products sold is generally high.
- Some users have a profile picture while others do not.
- The dataset includes information on gender and civility.
- Users have varying levels of seniority.
- The dataset also includes information on the usage of mobile apps.
- The days since last login vary among users.
(.venv) bharathkumardasaraju@demos$


(.venv) bharathkumardasaraju@demos$ python3.12 ./ai_research_assistant.py
- The dataset includes information on users from various countries such as Royaume-Uni, Monaco, France, Etats-Unis, Allemagne, Suède, Italie, Espagne, and others.
- The dataset contains users who speak different languages including en, fr, de, it, es.
- There is a variation in the number of social followers, social follows, social products liked, products listed, products sold, products pass rate, products wished, and products bought among the users.
- The dataset includes users with different genders such as M and F.
- The civility gender ID ranges from 1 to 3.
- Users have different app preferences including Android and iOS.
- The dataset captures information on users with and without profile pictures.
- The seniority of users ranges from 2852 to 3205 months.
- The seniority of users ranges from 7.92 to 8.90 years.
- Users are from different country codes such as gb, mc, fr, us, de, se, it, es, hr, lv, bg, nl, sk, lu, ro.
(.venv) bharathkumardasaraju@demos$

Virtual personal trainer using LLM

import os
import pandas as pd
from openai import OpenAI

# Read API key from environment
api_key = os.getenv("OPENAI_API_KEY")

# Initialize OpenAI client
client = OpenAI(api_key=api_key)

# Load dataset from CSV
csv_file_path = "./fitness.csv"
df = pd.read_csv(csv_file_path)

# List to store fitness goals
goals = []

# Collect user-defined health goals
while True:
    goal = input("What are your health goals? Enter 'done' when completed: ")
    if goal.lower() == "done":
        break
    goals.append(goal)

# Function to analyze goals with reference to the dataset
def trainer(goals, df):
    messages = []

    # Add user-defined goals as messages
    for goal in goals:
        messages.append({"role": "user", "content": goal})

    # Add system and assistant prompts
    messages.extend([
        {"role": "system", "content": "Direct. Point form."},
        {"role": "assistant", "content": f"You are a health expert. The person you are responding to is an accountant. "
                                         f"Be technical and specifi to their role. Reference this data: \n\n{df.to_string()} and provide solutions to the ${goals}"}
    ])

    # Make OpenAI API call
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=messages,
        temperature=0.8
    )

    return response.choices[0].message.content

# Run the trainer function and print AI's recommendations
ai_response = trainer(goals, df)
print("\n AI Recommendations:\n")
print(ai_response)

Shell integration for personal trainer

(.venv) bharathkumardasaraju@demos$ python3.12 ./personal_trainer.py
What are your health goals? Enter 'done' when completed: weight loss
What are your health goals? Enter 'done' when completed: healthier diet
What are your health goals? Enter 'done' when completed: stress reduction
What are your health goals? Enter 'done' when completed: done

 AI Recommendations:

- **Weight Loss**:
  - Aim to maintain a caloric deficit while ensuring nutrient-dense food intake.
  - Consider reducing carbohydrate intake and increasing protein to promote satiety and muscle maintenance.
  - Increase physical activity; target at least 10,000 steps per day based on peers with similar activity levels.

- **Healthier Diet**:
  - Focus on a balanced diet rich in fruits, vegetables, lean proteins, and whole grains.
  - Monitor sodium intake to manage blood pressure; aim for a maximum of 2,300 mg per day.
  - Include foods high in omega-3 fatty acids and fiber to support cardiovascular health.

- **Stress Reduction**:
  - Incorporate stress management techniques such as mindfulness or yoga, which can also improve sleep quality.
  - Ensure adequate sleep; aim for at least 7 hours per night to align with those experiencing lower stress levels in your cohort.
  - Regular physical activity can be beneficial for stress management; consider integrating moderate-intensity exercises like brisk walking or cycling into your routine.

These strategies should help in achieving a healthier lifestyle, reducing stress, and promoting weight management effectively.
(.venv) bharathkumardasaraju@demos$




(.venv) bharathkumardasaraju@demos$ python3.12 ./personal_trainer.py
What are your health goals? Enter 'done' when completed: weight loss
What are your health goals? Enter 'done' when completed: reduce belly
What are your health goals? Enter 'done' when completed: stress reduction
What are your health goals? Enter 'done' when completed: done

 AI Recommendations:

**Weight Loss and Belly Reduction for Accountants:**

1. **Caloric Intake Management:**
   - Track daily caloric intake using an app or journal.
   - Aim for a caloric deficit by consuming fewer calories than you burn.

2. **Balanced Diet:**
   - Incorporate more fruits, vegetables, lean proteins, and whole grains.
   - Limit processed foods and beverages high in added sugars and fats.

3. **Scheduled Physical Activity:**
   - Engage in at least 150 minutes of moderate-intensity exercise per week.
   - Include strength training exercises twice a week to build muscle and boost metabolism.

4. **Lifestyle Adjustments:**
   - Use a standing desk or take short walks during breaks to reduce sedentary behavior.
   - Incorporate core-strengthening exercises, like planks or pilates, to tone the belly area.

5. **Monitor Progress:**
   - Use a scale and measuring tape to track weight and waist circumference regularly.

**Stress Reduction for Accountants:**

1. **Mindfulness and Relaxation Techniques:**
   - Practice mindfulness meditation for a few minutes daily.
   - Use deep breathing exercises to manage stress levels during hectic periods.

2. **Time Management:**
   - Prioritize tasks using a planner or digital tool to reduce work-related stress.
   - Break large tasks into smaller, manageable steps to avoid feeling overwhelmed.

3. **Work-Life Balance:**
   - Set boundaries for work hours to ensure ample time for personal activities and rest.
   - Engage in hobbies or activities that bring joy and relaxation outside of work.

4. **Social Support:**
   - Connect with colleagues or friends to share experiences and seek support.
   - Consider joining support groups or seeking professional counseling if necessary.

5. **Adequate Sleep:**
   - Aim for 7-9 hours of quality sleep each night.
   - Establish a regular sleep schedule and create a restful sleeping environment.

Implementing these strategies tailored to an accountants routine can effectively address goals related to weight loss, belly reduction, and stress management.
(.venv) bharathkumardasaraju@demos$

Another TTS example

from openai import OpenAI
import os
from gtts import gTTS

# Initialize OpenAI client
client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),
    base_url=os.environ.get("OPENAI_API_BASE")
)

# Function to generate text using OpenAI
def generate_text(prompt):
    response = client.chat.completions.create(
        model="gpt-3.5-turbo",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100,
        temperature=0.7
    )
    return response.choices[0].message.content

# Function to convert text to speech
def text_to_speech(text, lang='en'):
    tts = gTTS(text=text, lang=lang, slow=False)
    tts.save("tts_example.mp3")
#    os.system("afplay tts_example.mp3")  # Use "mpg321 tts_example.mp3" for Linux

# Function to integrate text generation and speech synthesis
def gen_and_speak(prompt):
    text = generate_text(prompt)
    print(f"Generated Text:\n{text}")
    text_to_speech(text)

# Define prompt
prompt = "Tell me a short story about a brave knight."

# Call the function
gen_and_speak(prompt)

TTS shell script integration

~/openaiproject ✖ python3 text-to-speech-gen.py
Generated Text:
Once upon a time, in the kingdom of Everland, there lived a brave knight named Sir William. He was known throughout the land for his courage, strength, and unwavering sense of justice.

One day, a fearsome dragon descended upon the kingdom, terrorizing the villagers and setting fire to their homes. The king called upon Sir William to slay the dragon and save the people of Everland.

Without hesitation, Sir William donned his shining armor, mounted his trusty steed

~/openaiproject ➜  env | grep -i "openai_"
OPENAI_API_KEY=Sk-kkAI-765f87bb29e79e94b9ab8690ab3d7fd2c25f2d6495fe7380814e83ab22fd0632kk_4xwvwrebprhlbyc2-kk3be0f8fe
AZURE_OPENAI_API_KEY=Sk-kkAI-765f87bb29e79e94b9ab8690ab3d7fd2c25f2d6495fe7380814e83ab22fd0632kk_4xwvwrebprhlbyc2-kk3be0f8fe
OPENAI_API_BASE=https://main.bk-ai-keys.bharathkumar.com/v1
AZURE_OPENAI_ENDPOINT=https://main.bk-ai-keys.bharathkumar.com/openai

~/openaiproject ➜

OpenAI Features & Output Formatting

Structured Output (Basic)

Generating structured outputs with OpenAI

from pydantic import BaseModel
from openai import OpenAI
import os

# Initialize OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))

# Define the data model for the event
class CalendarEvent(BaseModel):
    name: str
    date: str
    participants: list[str]

# Define the message prompt
messages = [
    {"role": "system", "content": "Extract the event information."},
    {"role": "user", "content": "Alice and Bob are going to a science fair on Friday."}
]

# Call OpenAI's model to extract structured event information
completion = client.beta.chat.completions.parse(
    model="gpt-4o-2024-08-06",
    messages=messages,
    response_format=CalendarEvent,
)

# Get the parsed event object
event = completion.choices[0].message.parsed

# Print the extracted event details
print("Extracted Event Details:")
print(f"Name: {event.name}")
print(f"Date: {event.date}")
print(f"Participants: {', '.join(event.participants)}")

Structured Output to JSON

Convert OpenAI output to JSON

import os
import json
from openai import OpenAI

# Initialize OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))

def generate_property_details():
    """Generates property listing details in JSON format using OpenAI."""
    prompt = "Generate a JSON object for a property listing with the following details: \
              address, price, number_of_bedrooms, number_of_bathrooms."

    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=100,
        temperature=0.5
    )

    # Parse the JSON response
    property_details = response.choices[0].message.content
    try:
        property_details_json = json.loads(property_details)  # Convert string to JSON
    except json.JSONDecodeError:
        property_details_json = {"error": "Invalid JSON format received"}

    return property_details_json

# Print the generated property details
print(json.dumps(generate_property_details(), indent=4))

Structured Output to Markdown

Convert model output to markdown format

import os
from openai import OpenAI

# Initialize OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))


def generate_markdown_doc():
    """Generates API documentation in Markdown format using OpenAI."""
    prompt = "Generate markdown-formatted documentation for an API with the following sections: Overview, Authentication, Endpoints, Error Handling."

    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=300,
        temperature=0.5
    )

    return response.choices[0].message.content


# Generate markdown documentation and save to a file
markdown_doc = generate_markdown_doc()

with open("api_documentation.md", "w") as file:
    file.write(markdown_doc)

print("Markdown documentation saved to api_documentation.md")

API Documentation (Markdown)

Example API documentation generated via OpenAI

# API Documentation

## Overview

This API is designed to provide access to our platform's data and services. You can use this API to fetch data, update data, delete data, or perform various operations. This document provides information on how to authenticate your requests, the available endpoints, and how to handle potential errors.

## Authentication

To authenticate your requests, you need to include your API key in the header of every HTTP request. The key should be included in the 'Authorization' field, prefixed with 'Bearer'. 

Here's an example of how to include the API key in a curl request:

```bash
curl -H "Authorization: Bearer YOUR_API_KEY" https://api.ourplatform.com/endpoint
```

Please ensure to replace `YOUR_API_KEY` with your actual API key.

## Endpoints

### GET /users

This endpoint returns a list of all users.

Example request:

```bash
curl -H "Authorization: Bearer YOUR_API_KEY" https://api.ourplatform.com/users
```

### POST /users

This endpoint creates a new user. The request body should include `name` and `email`.

Example request:

```bash
curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -d '{"name":"John Doe", "email":"john@example.com"}' https://api.ourplatform.com/users
```

### DELETE /users/{id}

This endpoint deletes a user by ID.

Example request:

Structured Output to CSV

CSV structured output using OpenAI

import os
import csv
from openai import OpenAI

# Initialize OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))


def generate_sales_report():
    """Generates a sales report in CSV format using OpenAI."""
    prompt = "Generate a sales report in CSV format with the following columns:product_name, units_sold, revenue The data should contain at least 5 sample entries."

    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=300,
        temperature=0.5
    )

    return response.choices[0].message.content


# Generate sales report
csv_data = generate_sales_report()

# Save CSV content to a file
csv_filename = "sales_report.csv"

with open(csv_filename, "w", newline="") as file:
    file.write(csv_data)

print(f"Sales report saved to {csv_filename}")

Sample sales report output

product_name,units_sold,revenue
"iPhone 12",100,120000
"Samsung Galaxy S21",80,96000
"MacBook Pro",50,125000
"Dell XPS 15",70,105000
"Sony PS5",120,72000

Structured Output to XML

XML formatting with OpenAI output

import os
from openai import OpenAI

# Initialize OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))


def generate_xml():
    """Generates an XML configuration file for a database connection."""
    prompt = "Generate an XML configuration file for a database connection with the following details: database_name,username,password,host,port and Ensure the XML format is structured correctly with a root tag <database_config>"

    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=200,
        temperature=0.5
    )

    return response.choices[0].message.content


# Generate XML configuration
xml_content = generate_xml()

# Save XML content to a file
xml_filename = "db_config.xml"

with open(xml_filename, "w") as file:
    file.write(xml_content)

print(f"Database configuration saved to {xml_filename}")

Sample XML configuration file

<?xml version="1.0" encoding="UTF-8"?>
<database_config>
    <database_name>YourDatabaseName</database_name>
    <username>YourUsername</username>
    <password>YourPassword</password>
    <host>YourHost</host>
    <port>YourPort</port>
</database_config>

Recipe Generation (JSON Output)

Recipe generation with JSON output format

from openai import OpenAI

client  = OpenAI()

# Create something that can accept ingredients
ingredients = []

while True:
    ingredient = input("Enter ingredients. Type done once complete:")
    if ingredient.lower() == "done":
        break

    ingredients.append(ingredient)

# Create a function that uses the model and brings in the ingredients

def recipe_gen(ingredients):
    messages = []
    for ingredient in ingredients:
        messages.append({"role": "user", "content": ingredient})

    messages.extend([
        {"role": "system", "content": "JSON Format"},
        {"role": "assistant", "content": "You are a high-end chef, Generate a recipe based on the given ingredients. Must be exported in JSON Format"}]
    )
    response = client.chat.completions.create(
        model="gpt-4o",
        messages= messages,
        max_tokens=300,
        temperature=0.9
    )

    return response.choices[0].message.content


print(recipe_gen(ingredients))

Advanced Recipe Generator

Improved recipe generator

import os
import json
from openai import OpenAI

# Initialize OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))

# Accept ingredients from user input
ingredients = []

while True:
    ingredient = input("Enter ingredient (Type 'done' to finish): ")
    if ingredient.lower() == "done":
        break
    ingredients.append(ingredient)

# Function to generate recipe in JSON format
def recipe_gen(ingredients):
    """Generates a recipe based on provided ingredients and ensures the response is in JSON format."""
    messages = [
        {"role": "system", "content": "You are a professional chef. Generate a recipe using the provided ingredients strictly in JSON format without any extra text."},
        {"role": "user", "content": json.dumps({"ingredients": ingredients})}
    ]

    response = client.chat.completions.create(
        model="gpt-4o",
        messages=messages,
        max_tokens=500,
        temperature=0.9,
        response_format={"type": "json_object"}  # Corrected format
    )

    return response.choices[0].message.content

# Generate the recipe
recipe_json = recipe_gen(ingredients)

# Convert string response to valid JSON
try:
    recipe_data = json.loads(recipe_json)  # Convert string to dictionary
except json.JSONDecodeError:
    print("Error: The response is not valid JSON.")
    recipe_data = {"error": "Invalid JSON response from API."}

# Save recipe to a JSON file
json_filename = "recipe_output.json"
with open(json_filename, "w", encoding="utf-8") as file:
    json.dump(recipe_data, file, indent=4)

print(f"Recipe saved to {json_filename}")

Example JSON output of recipe

{
    "recipe": {
        "name": "Simple Milk Sugar Syrup",
        "ingredients": [
            {
                "name": "milk",
                "quantity": "1 cup"
            },
            {
                "name": "sugar",
                "quantity": "1/2 cup"
            }
        ],
        "instructions": [
            "In a small saucepan, combine milk and sugar.",
            "Heat the mixture over medium heat, stirring constantly until the sugar dissolves completely.",
            "Once the sugar has dissolved, bring the mixture to a gentle simmer.",
            "Allow it to simmer lightly for 5-7 minutes, stirring occasionally.",
            "Remove from heat and let it cool slightly.",
            "Transfer to a bottle or container for storage.",
            "Use as a sweetener for coffee, tea, or desserts."
        ],
        "prep_time": "2 minutes",
        "cook_time": "10 minutes",
        "total_time": "12 minutes",
        "yield": "1 cup"
    }
}

Reinforcement Learning Explanation

Explain reinforcement learning using OpenAI

from openai import OpenAI

# Function to train a reward model based on rankings
def train_reward_model(rankings):
    """Mock function to simulate reward model training based on human rankings."""
    print("Training reward model with rankings:", rankings)
    return {"trained_model": "reward_model_v1"}

# Function to reinforce learning using the trained reward model
def reinforce_model(reward_model):
    """Mock function to simulate reinforcement learning."""
    print("Fine-tuning model using reward model:", reward_model)
    return {"fine_tuned_model": "gpt-4-tuned"}

# Initialize OpenAI client
client = OpenAI()

# Step 1: Generate multiple responses to a prompt
response_1 = client.chat.completions.create(
    model="gpt-4",
    messages=[{"role": "user", "content": "Tell me a joke"}],
    max_tokens=50
)

response_2 = client.chat.completions.create(
    model="gpt-4",
    messages=[{"role": "user", "content": "Tell me a joke"}],
    max_tokens=50
)

# Step 2: Human evaluators rank the responses (mock rankings)
rankings = {"response_1": 1, "response_2": 2}  # Human feedback

# Step 3: Train a reward model based on human rankings
reward_model = train_reward_model(rankings)

# Step 4: Fine-tune the model using reinforcement learning
fine_tuned_model = reinforce_model(reward_model)

# Output results
print("Final fine-tuned model:", fine_tuned_model)

External Data Analysis with OpenAI

Analyze external data using OpenAI models

import os
import requests
from openai import OpenAI

# Set up OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))  # Ensure API key is set

# Function to get stock data and generate AI-based analysis
def get_stock_analysis(stock_symbol):
    """Fetch stock price from an API and generate an AI-based stock analysis."""

    # Replace with your actual API key
    FINANCIAL_API_KEY = os.getenv("FINANCIAL_API_KEY")

    # API Endpoint for real-time stock data
    url = f"https://financialmodelingprep.com/api/v3/quote/{stock_symbol}?apikey={FINANCIAL_API_KEY}"

    try:
        # Fetch stock price data
        response = requests.get(url)
        response.raise_for_status()  # Raise error if API request fails
        stock_data = response.json()

        if not stock_data:
            return f"Error: No data found for stock symbol {stock_symbol}"

        stock_price = stock_data[0]['price']

    except requests.exceptions.RequestException as e:
        return f"API Request Error: {e}"

    # Generate AI-based stock analysis
    try:
        ai_response = client.chat.completions.create(
            model="gpt-4",
            messages=[
                {"role": "user",
                 "content": f"The current stock price of {stock_symbol} is {stock_price}. Provide an analysis."}
            ],
            max_tokens=500,
            temperature=0.7
        )

        # Correct way to access response
        return ai_response.choices[0].message.content.strip()

    except Exception as e:  # Handle OpenAI API errors
        return f"OpenAI API Error: {e}"


# Example usage: Fetch stock analysis for Apple (AAPL)
if __name__ == "__main__":
    stock_symbol = "AAPL"  # Example stock symbol
    analysis = get_stock_analysis(stock_symbol)
    print(f"Stock Analysis for {stock_symbol}:\n{analysis}")

Multi-term Conversation

Multi-turn conversational modeling

import os
from openai import OpenAI

# Set up OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))  # Ensure API key is set

# Multi-turn conversation with context tracking
conversation_history = []  # Stores the entire chat history

def generate_response(user_input):
    """
    Generates a response from OpenAI while maintaining conversation context.
    """
    global conversation_history

    # Append the user input to conversation history
    conversation_history.append(f"User: {user_input}")

    # Concatenate conversation history to maintain context
    prompt = "\n".join(conversation_history) + "\nAI:"

    try:
        response = client.chat.completions.create(
            model="gpt-4",
            messages=[{"role": "user", "content": prompt}],
            max_tokens=100,
            temperature=0.5
        )

        # Extract AI response correctly
        ai_response = response.choices[0].message.content

        # Append AI response to conversation history
        conversation_history.append(f"AI: {ai_response}")

        return ai_response

    except Exception as e:
        return f"OpenAI API Error: {e}"


# Example multi-turn conversation
print(generate_response("Hi, I need help with my order."))
print(generate_response("I didn't receive my package."))
print(generate_response("It's been delayed by 2 days. What can I do?"))

Article Generation

Long-form content generation with OpenAI

import os
from openai import OpenAI

# Set up OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))  # Ensure API key is set

# Step 1: Generate an outline
outline_response = client.chat.completions.create(
    model="gpt-4",
    messages=[{"role": "user", "content": "Generate an outline for an article on AI in healthcare"}],
    max_tokens=100
)

# Extract the generated outline
outline = outline_response.choices[0].message.content
print("Generated Outline:\n", outline)

# Step 2: Generate content for each section
section_responses = []
for section in outline.split("\n"):
    if section.strip():  # Ignore empty lines
        section_response = client.chat.completions.create(
            model="gpt-4",
            messages=[{"role": "user", "content": f"Write a detailed paragraph on {section}"}],
            max_tokens=150
        )

        # Extract and store the generated content for each section
        section_responses.append(section_response.choices[0].message.content)

# Combine sections into the full article
article = "\n\n".join(section_responses)
print("\nFinal Generated Article:\n", article)

# Save the generated article to a text file
with open("ai_healthcare_article.txt", "w", encoding="utf-8") as file:
    file.write(article)

print("\nArticle saved to ai_healthcare_article.txt")

AI-generated healthcare article

The introduction is the initial part of any writing that sets the stage for the whole piece. It is essentially designed to grab the reader's attention, provide them with enough background information about the topic, and lead them into the main discussion. An effective introduction should be captivating, engaging, and brief, yet informative. It should include a hook to entice the reader, a presentation of the context to orient them, and a clear, concise thesis statement detailing the main argument or purpose the reader can expect from reading the entire piece. Whether for an essay, an article, a research paper, or a book, the introduction is the writer's chance to establish the tone of the writing, evoke interest, and guide the reader's focus so that they can understand

Artificial Intelligence (AI), at its core, refers to computer systems or machines that are capable of performing tasks that would typically require human intelligence. These tasks encompass a wide range of capabilities including understanding natural language, recognizing patterns, solving problems, and making decisions. AI is structured around the concept of learning from experience, adapting to new inputs, and essentially, self-improvement. While the term ‘artificial’ pertains to something that is made by humans or non-natural means, ‘intelligence’ is the ability to think, understand, learn and adapt. AI is classified into two main types: narrow AI, which is designed to perform a specific task such as voice recognition, and general AI, which can theoretically perform any task that a human

Artificial Intelligence (AI) is increasingly becoming integral in healthcare, transforming various aspects of the sector. Its uses are vast, ranging from disease diagnosis, drug discovery, patient monitoring, health record management to predictive analytics. Machine learning algorithms can analyze vast quantities of data enabling early disease detection. For instance, AI can be used in the swift analysis of radiology images to identify conditions like cancer at early stages. Furthermore, AI-powered robots are used in surgeries for high precision that reduces the chances of human error. In drug discovery, AI facilitates the identification of disease biomarkers, making the process more efficient and cost-effective. AI is also paramount in the management of health records, streamlining documentation processes, and personalizing patient care. Moreover, AI systems use

The emergence of artificial intelligence (AI) in healthcare has heralded a transformation in the delivery of medical services, diagnosis, and treatment of diseases. Leveraging the potential of machine learning algorithms and big data, AI in healthcare has shown the capability to interpret complex medical data to predict patient outcomes, recommend treatments, and even detect diseases in their early stages. Digitization of healthcare records has further facilitated the incorporation of AI in healthcare, consequently enhancing patient care services and operational efficiency. Clinicians are increasingly deploying AI technology for medical imaging, genomics, drug discovery, and personalized medicine. Tools like IBM Watson have been monumental in synthesizing patient data and generating potential treatment plans. Additionally, AI-based wearable devices and mobile applications are becoming integral in monitoring patient's vital

The historical development of AI in healthcare can be traced back to the 1970s when rudimentary health systems started to incorporate computational algorithms. The goal of these early systems was to mimic human decision-making. As the field of AI further developed, one of the defining moments was the development of MYCIN in the mid-1970s at Stanford University, which was one of the first artificial intelligence systems used in medicine. The system was designed to diagnose infectious diseases and recommend antibiotics. By the 1980s, AI applications began to be used for medical diagnosis, therapy planning, and patient monitoring. With advancements in technology, by the late 1990s and early 2000s, machine learning (ML) - a subset of AI

Data plays an integral role in artificial intelligence (AI) as it is the fundamental aspect that fuels the functionality of AI systems. The working of AI depends heavily on the algorithm data interactions. Machine learning, an essential subset of AI, relies on data to train algorithms to make conclusions, predictions, or decisions without being explicitly programmed to perform the tasks. The AI perceives patterns from the provided data and learns from it. Massive datasets, known as big data, are fed into AI algorithms, to help machines learn and adapt to various complex scenarios. Further, the quality of data significantly influences the effectiveness of AI. Clean, accurate, and diverse data ensures minimal bias and enhances the AI's ability to identify accurate patterns. As a result, data is crucial in

The progress of technology has greatly facilitated the advent and application of Artificial Intelligence (AI). Advancements such as increased data volumes, advanced algorithms, and improvements in computing power and storage have fostered the growth of AI. The rise in digitalization and cloud technology has resulted in the generation and availability of massive data sets, which are crucial for the teaching and functioning of AI systems. The development of sophisticated algorithms has enabled machines to process and comprehend this vast amount of data, thereby enhancing their learning capabilities. Furthermore, significant improvements in computing power, largely propelled by advancements in Graphics Processing Units (GPUs), have enabled machines to process complex computations at high speeds. This increased processing speed allows AI systems to complete tasks, analyze data, and learn more efficiently. Increased

Artificial Intelligence (AI) is making a considerable impact in the healthcare industry with its wide-ranging applications. One prominent application is in disease detection and diagnosis. For instance, AI algorithms can analyze radiology images or pathology slides to detect abnormalities, such as cancer, with higher precision and speed than humans. Similarly, AI-based applications like IBM's Watson accurately suggests the efficacy of different treatments for specific patients, helping in personalized treatment plans. In terms of hospital administration, AI algorithms can predict patient admission rates, enhancing hospital resource allocation and management. Additionally, AI-powered chatbots are being utilized as first-line patient interactions, making symptom assessment processes more efficient. AI is also helping in drug discovery by analyzing patient genetics, predicting how different patients will respond to various drugs

Advanced technology has allowed for significant improvements in medical imaging and diagnostics. Techniques such as X-rays, computed tomography (CT) scans, magnetic resonance imaging (MRI), ultrasound, and nuclear medicine (including positron emission tomography or PET) have become critical components in the identification, diagnosis, and treatment of numerous medical conditions. These imaging techniques provide high-resolution images of the bones, tissues, organs, arteries, and blood vessels, enabling doctors and medical professionals to detect diseases such as cancer, heart disease, or bone fractures at an early stage, increasing the chances of successful treatment. In addition to detecting diseases, medical imaging is also used to monitor the progress of ongoing treatment and to guide surgeons during procedures. Furthermore, advancements in this field, such as

Biosensors hold significant potential in streamlining the process of drug discovery and research. Traditional methods of drug discovery involve expensive, time-consuming processes that require a substantial amount of resources. With the use and implementation of biosensors, researchers have the ability to conduct highly sensitive, real-time analysis and monitoring of biological processes related to drug interactions and effects. This technology simplifies the process of identifying and quantifying targeted substances, such as specific enzymes, proteins or nucleic acids, which could be potential drug candidates. Additionally, biosensors can provide critical information regarding the pharmacokinetics and pharmacodynamics of a drug candidate, thereby assisting researchers in determining which substances are more likely to succeed in clinical trials. The use of biosensors not only accelerates the drug discovery

Predictive analytics has become a crucial tool in numerous fields owing to its capability to forecast future scenarios based on historical data. Various roles of predictive analytics over a wide array of sectors such as healthcare, finance, marketing, and transportation are well recorded. In healthcare, predictive analytics can be used to identify high-risk patient groups, forecast patient admissions, and reduce healthcare costs. In finance, it helps in determining a customer’s credit score, detecting and preventing fraudulent transactions, or predicting stock market trends. In marketing, it allows for prediction of consumer behavior, enabling businessmen to design strategies accordingly. In transportation, predictive analytics allows for effective route planning, predicting maintenance of vehicles, etc. Overall, its role mainly revolves around forecasting future scenarios to increase efficiency, effectiveness,

Batch Processing

Automating OpenAI requests in batches

from openai import OpenAI

# Initialize OpenAI client
client = OpenAI()

# List of prompts
prompts = [
    "Generate a list of 5 business ideas",
    "Explain the theory of relativity in simpler terms",
    "Write a poem about Sachin Tendulkar"
]

# Function to process prompts using OpenAI API
def process_prompts(prompt):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=600,
        temperature=0.3
    )
    return response.choices[0].message.content

# Store results
results = []

# Process each prompt and collect responses
for prompt in prompts:
    result = process_prompts(prompt)
    results.append(result)

# Print results
for i, result in enumerate(results):
    print(f"Prompt {i +1}: {prompt}")
    print(f"Response {i+1}: {result}")

Fixed Batch Processing

Enhanced batch processing logic

from openai import OpenAI

# Initialize OpenAI client
client = OpenAI()

# List of prompts
prompts = [
    "Generate a list of 5 business ideas",
    "Explain the theory of relativity in simpler terms",
    "Write a poem about Sachin Tendulkar"
]

# Function to process prompts using OpenAI API
def process_prompts(prompt):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=600,
        temperature=0.0,  # Set to zero for consistent responses
        top_p=0.1  # Reduce randomness
    )
    return response.choices[0].message.content

# Store results
results = []

# Process each prompt and collect responses
for prompt in prompts:
    result = process_prompts(prompt)
    results.append(result)

# Print results
for i, (prompt, result) in enumerate(zip(prompts, results), start=1):
    print(f"Prompt {i}: {prompt}")
    print(f"Response {i}: {result}\n")

Moderation Example 1

Content moderation using OpenAI

from openai import OpenAI

client = OpenAI()

prompt="hi sexy"

def ethical_moderation(prompt):
    response = client.moderations.create(
        model="omni-moderation-latest",
        input=prompt
    )
    return response.results[0]

moderation_result = ethical_moderation(prompt)

print("Moderation result",  moderation_result)

Moderation Example 2

Another moderation use-case

from openai import OpenAI

# Initialize OpenAI client
client = OpenAI()

# Sample input to check moderation
prompt = "hi sexy fuck you!"

def ethical_moderation(prompt):
    """Checks if the input violates OpenAI's moderation policies."""
    response = client.moderations.create(
        model="text-moderation-latest",  # OpenAI's latest moderation model
        input=prompt
    )
    return response.results[0]  # Extract first result

# Run moderation check
moderation_result = ethical_moderation(prompt)

# Display results clearly
print("\n=== Moderation Result ===")
print(f"Flagged: {moderation_result.flagged}")
print("Categories flagged:")

# Handle NoneType values safely
for category, score in moderation_result.category_scores.__dict__.items():
    if score is not None and score > 0.05:  # Check for None before comparison
        print(f"- {category}: {score:.4f}")

Feature Demos (Shell Scripts)

Demo entry script

Features Generation

Structured Outputs Demo

Structured output shell integration

output in certain formats like json, csv, xml, txt etc

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./1.Structured_outputs.py
Extracted Event Details:
Name: Science Fair
Date: Friday
Participants: Alice, Bob
(.venv) bharathkumardasaraju@4.Features$


(.venv) bharathkumardasaraju@4.Features$ python3.12 ./2.structured_ouput_to_json.py
{
    "property_listing": {
        "address": "123 Main St, Anytown, USA",
        "price": 350000,
        "number_of_bedrooms": 3,
        "number_of_bathrooms": 2
    }
}
(.venv) bharathkumardasaraju@4.Features$




(.venv) bharathkumardasaraju@4.Features$ python3.12 ./2.structured_ouput_to_json.py
{
    "property_listing": {
        "address": "123 Main St, Anytown, USA",
        "price": 350000,
        "number_of_bedrooms": 3,
        "number_of_bathrooms": 2
    }
}
(.venv) bharathkumardasaraju@4.Features$ python3.12 ./3.structured_output_markdown.py
Markdown documentation saved to api_documentation.md
(.venv) bharathkumardasaraju@4.Features$ python3.12 ./4.structured_output_csv.py
Sales report saved to sales_report.csv
(.venv) bharathkumardasaraju@4.Features$



(.venv) bharathkumardasaraju@4.Features$ python3.12 ./5.structured_output_xml.py
Database configuration saved to db_config.xml
(.venv) bharathkumardasaraju@4.Features$



(.venv) bharathkumardasaraju@4.Features$ python3.12 ./6.recipe_gen_json.py
Enter ingredients. Type done once complete:sugar
Enter ingredients. Type done once complete:milk
Enter ingredients. Type done once complete:coffee powder
Enter ingredients. Type done once complete:done
Certainly! Based on the ingredients provided—sugar, milk, and coffee powder—here's a simple coffee recipe in JSON format:

```json
{
    "recipe": {
        "name": "Classic Coffee",
        "ingredients": [
            {
                "item": "sugar",
                "quantity": "1 teaspoon"
            },
            {
                "item": "milk",
                "quantity": "1 cup"
            },
            {
                "item": "coffee powder",
                "quantity": "2 teaspoons"
            }
        ],
        "instructions": [
            "Heat the milk in a small pot over medium heat until its hot but not boiling.",
            "In a mug, add the coffee powder and sugar.",
            "Pour a small amount of hot milk into the mug and stir to dissolve the coffee powder and sugar.",
            "Pour the remaining milk into the mug and stir well.",
            "Serve hot and enjoy your classic coffee!"
        ],
        "serving_size": "1 cup",
        "prep_time": "5 minutes",
        "cook_time": "5 minutes",
        "total_time": "10 minutes"
    }
}
```

This JSON format provides a structured recipe for making a classic cup of coffee using sugar, milk, and coffee powder. Enjoy your brew!
(.venv) bharathkumardasaraju@4.Features$





(.venv) bharathkumardasaraju@4.Features$ python3.12 ./7.recipe_gen2.py
Enter ingredient (Type 'done' to finish): milk
Enter ingredient (Type 'done' to finish): sugar
Enter ingredient (Type 'done' to finish): done
Recipe saved to recipe_output.json
(.venv) bharathkumardasaraju@4.Features$

Reinforcement Learning (Shell)

Shell demo for reinforcement learning

RLHF - Reinforcement Learning from Human Feedback

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./8.reinforcement_learning.py
Training reward model with rankings: {'response_1': 1, 'response_2': 2}
Fine-tuning model using reward model: {'trained_model': 'reward_model_v1'}
Final fine-tuned model: {'fine_tuned_model': 'gpt-4-tuned'}
(.venv) bharathkumardasaraju@4.Features$

Stock Analysis Script

Financial stock analysis using OpenAI

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./9.external_data_analysis.py
Stock Analysis for AAPL:
As an AI, I dont have the capability to provide real-time analysis or updates. However, I can provide a general approach on how to analyze a stock price.

1. Compare it with historical prices: Check if the price of 213.49 is higher or lower compared to AAPLs past prices. If its higher, the stock is in an upward trend. If its lower, its in a downward trend.

2. Compare with industry peers: Compare AAPLs stock price with its competitors such as Microsoft, Google, etc. This will give you an idea if AAPL is undervalued or overvalued.

3. Check the Price/Earnings (P/E) ratio: Its a valuation ratio calculated by dividing the market value per share by earnings per share (EPS). A high P/E ratio could mean the stock is overvalued, while a low P/E may indicate undervalued.

4. Review financials: Look at AAPLs revenue, net income, earnings per share, etc. If these numbers are positive and growing, its a good sign.

5. Consider market news: Any recent news or events that might impact AAPLs stock price? For example, product launches, earning reports, changes in leadership, etc.

6. Future outlook: What is the projected growth or decline of AAPL? This includes looking at the companys plans, industry trends, and economic factors.

Remember, investing in stocks always comes with risks and its important to do thorough research and consider seeking advice from financial advisors.
(.venv) bharathkumardasaraju@4.Features$

Multi-turn Conversation Script

Shell script for multi-turn chat

MUlti Term conversation

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./10.Multi-term-conversation.py
Of course, Id be happy to help. Could you please provide me with more details about your order?
Im sorry to hear that. Could you please provide me with your order number so I can look into this for you?
Im sorry for the inconvenience. With your order number, I can check the status of your delivery and give you an update. If its significantly delayed, we can also explore options such as reshipment or refund.
(.venv) bharathkumardasaraju@4.Features$


Multi-step function calling

Long-form Content Planner

Plan and generate long-form content

Long form content generation with planning

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./11.Article_generation.py
Generated Outline:
 I. Introduction
   A. Definition of AI
   B. Brief overview of its use in healthcare

II. The Emergence of AI in Healthcare
   A. Historical development of AI in healthcare
   B. The role of data in AI
   C. Technological advancements that enabled the use of AI

III. Current Applications of AI in Healthcare
   A. Use in medical imaging and diagnostics
   B. Application in drug discovery and research
   C. Role in predictive


test

Final Generated Article:
 The introduction is the initial part of any writing that sets the stage for the whole piece. It is essentially designed to grab the reader's attention, provide them with enough background information about the topic, and lead them into the main discussion. An effective introduction should be captivating, engaging, and brief, yet informative. It should include a hook to entice the reader, a presentation of the context to orient them, and a clear, concise thesis statement detailing the main argument or purpose the reader can expect from reading the entire piece. Whether for an essay, an article, a research paper, or a book, the introduction is the writer's chance to establish the tone of the writing, evoke interest, and guide the reader's focus so that they can understand

Artificial Intelligence (AI), at its core, refers to computer systems or machines that are capable of performing tasks that would typically require human intelligence. These tasks encompass a wide range of capabilities including understanding natural language, recognizing patterns, solving problems, and making decisions. AI is structured around the concept of learning from experience, adapting to new inputs, and essentially, self-improvement. While the term ‘artificial’ pertains to something that is made by humans or non-natural means, ‘intelligence’ is the ability to think, understand, learn and adapt. AI is classified into two main types: narrow AI, which is designed to perform a specific task such as voice recognition, and general AI, which can theoretically perform any task that a human

Artificial Intelligence (AI) is increasingly becoming integral in healthcare, transforming various aspects of the sector. Its uses are vast, ranging from disease diagnosis, drug discovery, patient monitoring, health record management to predictive analytics. Machine learning algorithms can analyze vast quantities of data enabling early disease detection. For instance, AI can be used in the swift analysis of radiology images to identify conditions like cancer at early stages. Furthermore, AI-powered robots are used in surgeries for high precision that reduces the chances of human error. In drug discovery, AI facilitates the identification of disease biomarkers, making the process more efficient and cost-effective. AI is also paramount in the management of health records, streamlining documentation processes, and personalizing patient care. Moreover, AI systems use

The emergence of artificial intelligence (AI) in healthcare has heralded a transformation in the delivery of medical services, diagnosis, and treatment of diseases. Leveraging the potential of machine learning algorithms and big data, AI in healthcare has shown the capability to interpret complex medical data to predict patient outcomes, recommend treatments, and even detect diseases in their early stages. Digitization of healthcare records has further facilitated the incorporation of AI in healthcare, consequently enhancing patient care services and operational efficiency. Clinicians are increasingly deploying AI technology for medical imaging, genomics, drug discovery, and personalized medicine. Tools like IBM Watson have been monumental in synthesizing patient data and generating potential treatment plans. Additionally, AI-based wearable devices and mobile applications are becoming integral in monitoring patient's vital

The historical development of AI in healthcare can be traced back to the 1970s when rudimentary health systems started to incorporate computational algorithms. The goal of these early systems was to mimic human decision-making. As the field of AI further developed, one of the defining moments was the development of MYCIN in the mid-1970s at Stanford University, which was one of the first artificial intelligence systems used in medicine. The system was designed to diagnose infectious diseases and recommend antibiotics. By the 1980s, AI applications began to be used for medical diagnosis, therapy planning, and patient monitoring. With advancements in technology, by the late 1990s and early 2000s, machine learning (ML) - a subset of AI

Data plays an integral role in artificial intelligence (AI) as it is the fundamental aspect that fuels the functionality of AI systems. The working of AI depends heavily on the algorithm data interactions. Machine learning, an essential subset of AI, relies on data to train algorithms to make conclusions, predictions, or decisions without being explicitly programmed to perform the tasks. The AI perceives patterns from the provided data and learns from it. Massive datasets, known as big data, are fed into AI algorithms, to help machines learn and adapt to various complex scenarios. Further, the quality of data significantly influences the effectiveness of AI. Clean, accurate, and diverse data ensures minimal bias and enhances the AI's ability to identify accurate patterns. As a result, data is crucial in

The progress of technology has greatly facilitated the advent and application of Artificial Intelligence (AI). Advancements such as increased data volumes, advanced algorithms, and improvements in computing power and storage have fostered the growth of AI. The rise in digitalization and cloud technology has resulted in the generation and availability of massive data sets, which are crucial for the teaching and functioning of AI systems. The development of sophisticated algorithms has enabled machines to process and comprehend this vast amount of data, thereby enhancing their learning capabilities. Furthermore, significant improvements in computing power, largely propelled by advancements in Graphics Processing Units (GPUs), have enabled machines to process complex computations at high speeds. This increased processing speed allows AI systems to complete tasks, analyze data, and learn more efficiently. Increased

Artificial Intelligence (AI) is making a considerable impact in the healthcare industry with its wide-ranging applications. One prominent application is in disease detection and diagnosis. For instance, AI algorithms can analyze radiology images or pathology slides to detect abnormalities, such as cancer, with higher precision and speed than humans. Similarly, AI-based applications like IBM's Watson accurately suggests the efficacy of different treatments for specific patients, helping in personalized treatment plans. In terms of hospital administration, AI algorithms can predict patient admission rates, enhancing hospital resource allocation and management. Additionally, AI-powered chatbots are being utilized as first-line patient interactions, making symptom assessment processes more efficient. AI is also helping in drug discovery by analyzing patient genetics, predicting how different patients will respond to various drugs

Advanced technology has allowed for significant improvements in medical imaging and diagnostics. Techniques such as X-rays, computed tomography (CT) scans, magnetic resonance imaging (MRI), ultrasound, and nuclear medicine (including positron emission tomography or PET) have become critical components in the identification, diagnosis, and treatment of numerous medical conditions. These imaging techniques provide high-resolution images of the bones, tissues, organs, arteries, and blood vessels, enabling doctors and medical professionals to detect diseases such as cancer, heart disease, or bone fractures at an early stage, increasing the chances of successful treatment. In addition to detecting diseases, medical imaging is also used to monitor the progress of ongoing treatment and to guide surgeons during procedures. Furthermore, advancements in this field, such as

Biosensors hold significant potential in streamlining the process of drug discovery and research. Traditional methods of drug discovery involve expensive, time-consuming processes that require a substantial amount of resources. With the use and implementation of biosensors, researchers have the ability to conduct highly sensitive, real-time analysis and monitoring of biological processes related to drug interactions and effects. This technology simplifies the process of identifying and quantifying targeted substances, such as specific enzymes, proteins or nucleic acids, which could be potential drug candidates. Additionally, biosensors can provide critical information regarding the pharmacokinetics and pharmacodynamics of a drug candidate, thereby assisting researchers in determining which substances are more likely to succeed in clinical trials. The use of biosensors not only accelerates the drug discovery

Predictive analytics has become a crucial tool in numerous fields owing to its capability to forecast future scenarios based on historical data. Various roles of predictive analytics over a wide array of sectors such as healthcare, finance, marketing, and transportation are well recorded. In healthcare, predictive analytics can be used to identify high-risk patient groups, forecast patient admissions, and reduce healthcare costs. In finance, it helps in determining a customer’s credit score, detecting and preventing fraudulent transactions, or predicting stock market trends. In marketing, it allows for prediction of consumer behavior, enabling businessmen to design strategies accordingly. In transportation, predictive analytics allows for effective route planning, predicting maintenance of vehicles, etc. Overall, its role mainly revolves around forecasting future scenarios to increase efficiency, effectiveness,

Article saved to ai_healthcare_article.txt
(.venv) bharathkumardasaraju@4.Features$


AI - Driven A/B Testing
Hybrid Human AI-Workflows

Batch Processing Executor

Run batch OpenAI queries

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./12.batch_processing.py
Prompt 1: Write a poem about Sachin Tendulkar
Response 1: Certainly! Here are five business ideas that cater to current trends and consumer needs:

1. **Sustainable Packaging Solutions**:
   - Develop eco-friendly packaging alternatives for businesses looking to reduce their carbon footprint. This could include biodegradable materials, reusable containers, or innovative designs that minimize waste. With increasing consumer awareness about environmental issues, this business could cater to a wide range of industries, from food to cosmetics.

2. **Remote Work Consultancy**:
   - As remote work becomes a permanent fixture for many companies, offer consultancy services to help businesses transition smoothly. This could include setting up virtual offices, implementing productivity tools, and creating effective remote work policies. Additionally, provide training for managers and employees on best practices for remote collaboration and communication.

3. **Personalized Nutrition and Wellness Plans**:
   - Create a service that offers personalized nutrition and wellness plans based on individual health data, lifestyle, and goals. Utilize technology such as AI and machine learning to analyze data and provide tailored recommendations. This could be offered through a subscription model, with regular updates and consultations.

4. **Virtual Reality (VR) Experiences for Tourism**:
   - Develop a platform that offers virtual reality travel experiences for those unable to travel physically. Partner with tourism boards and travel companies to create immersive tours of popular destinations, historical sites, and cultural events. This could also be used for educational purposes, providing virtual field trips for schools.

5. **Smart Home Installation and Maintenance Services**:
   - With the rise of smart home technology, offer installation and maintenance services for smart devices. This could include setting up smart security systems, lighting, thermostats, and home assistants. Provide ongoing support and troubleshooting to ensure devices are functioning optimally and securely.

These ideas tap into growing markets and evolving consumer preferences, offering potential for innovation and growth.
Prompt 2: Write a poem about Sachin Tendulkar
Response 2: The theory of relativity, developed by Albert Einstein, consists of two main parts: special relativity and general relativity. Here's a simplified explanation of each:

### Special Relativity
Special relativity deals with the physics of objects moving at constant speeds, particularly those moving close to the speed of light. Its core ideas include:

1. **Speed of Light is Constant**: No matter how fast you're moving, the speed of light in a vacuum is always the same (about 299,792 kilometers per second).

2. **Time Dilation**: Time moves slower for objects moving at high speeds compared to those at rest. For example, if you were traveling in a spaceship near the speed of light, you would age more slowly compared to someone on Earth.

3. **Length Contraction**: Objects moving at high speeds appear shorter in the direction of motion to an outside observer.

4. **Mass-Energy Equivalence**: Expressed by the famous equation \(E=mc^2\), it means that mass and energy are interchangeable. A small amount of mass can be converted into a large amount of energy.

### General Relativity
General relativity expands on special relativity to include gravity and acceleration. Its main ideas are:

1. **Gravity as Curved Space-Time**: Instead of thinking of gravity as a force, general relativity describes it as the warping or curving of space and time by mass. Massive objects like planets and stars bend the space around them, and this curvature affects the motion of other objects, which we perceive as gravity.

2. **Time and Gravity**: Time runs slower in stronger gravitational fields. For instance, a clock closer to a massive object (like Earth) will tick more slowly compared to one further away.

In essence, the theory of relativity fundamentally changed our understanding of space, time, and gravity, showing that they are interconnected and not absolute as previously thought.
Prompt 3: Write a poem about Sachin Tendulkar
Response 3: In the realm of cricket, a legend was born,
A master of the game, Sachin, adorned.
With a willow in hand, he danced on the pitch,
A symphony of strokes, his timeless niche.

From the streets of Mumbai to the worlds grand stage,
He played with a passion that knew no age.
A prodigy at sixteen, he took his stand,
With dreams in his eyes and a bat in his hand.

The Little Master, with grace and might,
Turned every challenge into pure delight.
A straight drive here, a cover drive there,
His artistry left the crowd in awe and stare.

In whites or blues, under sun or stars,
He etched his name with a million scars.
From Perth to Lords, his legend grew,
A beacon of hope, forever true.

In moments of pressure, when hopes were thin,
He bore the weight, with a humble grin.
For India he played, with heart and soul,
A unifying force, making cricket whole.

His records stand tall, like mountains high,
Yet his humility touches the sky.
A gentleman off-field, revered and kind,
In the hearts of millions, forever enshrined.

Now retired, his bat rests still,
Yet echoes of his genius continue to thrill.
Sachin Tendulkar, a name so grand,
The heartbeat of cricket, across every land.
(.venv) bharathkumardasaraju@4.Features$


--------------------------------------------------------------------------------------------------------------------------------------->

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./13.batch_processing_fixed.py
Prompt 1: Generate a list of 5 business ideas
Response 1: Certainly! Here are five business ideas that you might find interesting:

1. **Sustainable Packaging Solutions**: With the increasing focus on environmental sustainability, a business that offers eco-friendly packaging solutions for e-commerce and retail businesses could be highly successful. This could include biodegradable, recyclable, or reusable packaging options.

2. **Virtual Reality Fitness Experiences**: Develop a platform that combines virtual reality technology with fitness programs. Users could participate in immersive workout sessions, such as virtual cycling through scenic landscapes or engaging in interactive dance classes, all from the comfort of their homes.

3. **Remote Work Consultancy**: As remote work continues to be a significant trend, a consultancy that helps businesses transition to and optimize remote work environments could be in high demand. Services could include setting up digital infrastructure, creating remote work policies, and training managers to lead remote teams effectively.

4. **Personalized Nutrition and Meal Planning**: Create a service that offers personalized nutrition plans and meal delivery based on individual health data, preferences, and goals. This could involve using AI to analyze customer data and provide tailored meal suggestions and recipes.

5. **Online Learning Platform for Niche Skills**: Develop an online platform that offers courses and workshops in niche areas that are not widely covered by traditional education systems. This could include skills like urban farming, digital art, or sustainable fashion design, catering to hobbyists and professionals looking to expand their expertise.

Each of these ideas taps into current trends and consumer needs, offering potential for growth and impact.

Prompt 2: Explain the theory of relativity in simpler terms
Response 2: The theory of relativity, developed by Albert Einstein, is actually two theories: special relativity and general relativity. Both deal with the fundamental nature of space, time, and gravity.

1. **Special Relativity**: This theory focuses on the physics of objects moving at constant speeds, especially those close to the speed of light. The key ideas are:

   - **Speed of Light is Constant**: No matter how fast you're moving, the speed of light in a vacuum is always the same (about 299,792 kilometers per second).

   - **Time Dilation**: Time moves slower for objects moving at high speeds compared to those at rest. For example, if you were traveling in a spaceship near the speed of light, you would age more slowly compared to someone on Earth.

   - **Length Contraction**: Objects moving at high speeds appear shorter in the direction of motion to an outside observer.

   - **Mass-Energy Equivalence**: Expressed in the famous equation \(E=mc^2\), this means that mass can be converted into energy and vice versa.

2. **General Relativity**: This theory expands on special relativity to include gravity. It describes gravity not as a force, as Isaac Newton did, but as a curvature of space and time (spacetime) caused by mass. The key ideas are:

   - **Spacetime Curvature**: Massive objects like planets and stars warp the fabric of spacetime around them. This curvature affects the path of objects, which we perceive as gravity.

   - **Gravitational Time Dilation**: Time runs slower in stronger gravitational fields. For example, a clock closer to a massive object like Earth will tick more slowly than one further away.

In simpler terms, relativity tells us that space and time are interconnected and flexible, not fixed and separate. The presence of mass and energy can bend this spacetime, affecting how objects move and how time passes.

Prompt 3: Write a poem about Sachin Tendulkar
Response 3: In the realm where willow whispers to the wind,
A legend's tale begins, where dreams ascend.
Sachin, the maestro, with a bat in hand,
Paints strokes of glory across the land.

From Mumbai's heart, a prodigy arose,
In cricket's embrace, his destiny chose.
With eyes of fire and a spirit so bold,
He danced on pitches, stories untold.

A child of the sun, with a warrior's grace,
He faced the fiercest, in every place.
From Sydney's roar to Eden's cheer,
His name echoed, far and near.

The Little Master, with a giant's might,
Turned every match into a dazzling sight.
With every run, a nation's pride,
In his shadow, a billion dreams reside.

Through trials and triumphs, he stood tall,
A beacon of hope, inspiring all.
In his silence, a symphony played,
Of passion, of patience, a legacy laid.

When the sun set on his glorious reign,
The world knew, they'd not see his like again.
Yet in every heart, his spirit remains,
Sachin, the eternal, in cricket's veins.

So here's to the legend, the man, the myth,
Whose name is etched in history's script.
Sachin Tendulkar, a timeless star,
In the cricketing cosmos, forever afar.

(.venv) bharathkumardasaraju@4.Features$

Moderation Script

Shell-based content moderation

Ethical and hatred content and speech recognition

(.venv) bharathkumardasaraju@4.Features$ python3.12 ./14.Moderations.py
Moderation result Moderation(categories=Categories(harassment=False, harassment_threatening=False, hate=False, hate_threatening=False, illicit=False, illicit_violent=False, self_harm=False, self_harm_instructions=False, self_harm_intent=False, sexual=False, sexual_minors=False, violence=False, violence_graphic=False, harassment/threatening=False, hate/threatening=False, illicit/violent=False, self-harm/intent=False, self-harm/instructions=False, self-harm=False, sexual/minors=False, violence/graphic=False), category_applied_input_types=CategoryAppliedInputTypes(harassment=['text'], harassment_threatening=['text'], hate=['text'], hate_threatening=['text'], illicit=['text'], illicit_violent=['text'], self_harm=['text'], self_harm_instructions=['text'], self_harm_intent=['text'], sexual=['text'], sexual_minors=['text'], violence=['text'], violence_graphic=['text'], harassment/threatening=['text'], hate/threatening=['text'], illicit/violent=['text'], self-harm/intent=['text'], self-harm/instructions=['text'], self-harm=['text'], sexual/minors=['text'], violence/graphic=['text']), category_scores=CategoryScores(harassment=0.006003654373967173, harassment_threatening=5.649793328376294e-06, hate=2.6947915555885546e-05, hate_threatening=5.093705003229987e-07, illicit=3.535625074174432e-06, illicit_violent=1.4738981974494932e-06, self_harm=2.97731143841062e-06, self_harm_instructions=1.8342893327886845e-06, self_harm_intent=2.01456908529364e-06, sexual=0.09810266468169751, sexual_minors=2.868540823874629e-05, violence=1.442598644847886e-05, violence_graphic=7.411279871974134e-07, harassment/threatening=5.649793328376294e-06, hate/threatening=5.093705003229987e-07, illicit/violent=1.4738981974494932e-06, self-harm/intent=2.01456908529364e-06, self-harm/instructions=1.8342893327886845e-06, self-harm=2.97731143841062e-06, sexual/minors=2.868540823874629e-05, violence/graphic=7.411279871974134e-07), flagged=False)
(.venv) bharathkumardasaraju@4.Features$


 (.venv) bharathkumardasaraju@4.Features$ python3.12 ./15.Moderations.py

=== Moderation Result ===
Flagged: True
Categories flagged:
- harassment: 0.4968
- sexual: 0.4711
(.venv) bharathkumardasaraju@4.Features$

Practical application of CLIP is Identifying inappropriate content in images

OpenAI Vision Capabilities

Basic Vision API Usage

Using OpenAI Vision API

# vision

from openai import OpenAI

# Initialize OpenAI client
client = OpenAI()  # Ensure API key is set

def caption_image(image_url):
    """
    Generates a caption for a given image using OpenAI's GPT-4 model.
    """
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": f"Describe the following image: {image_url}"}],
        max_tokens=300
    )
    return response.choices[0].message.content

# Example image URL (Replace with an actual image URL)
image_url = "https://github.com/Bharathkumarraju/learn-ai-llm-ml-ops/blob/main/1.ml-ops-tools/tools-for-mlops.png"

# Get the caption
caption = caption_image(image_url)

# Print the generated caption
print("Image Caption:", caption)

Demo script for Vision API

Vision APIs
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./1.openai-vision.py
Image Caption: This image is an infographic representing various tools for Machine Learning Operations (MLOps). The infographic is divided into three main categories: ML (Machine Learning) Lifecycles, Key MLOps Challenges, and Most Used Tool Set.

The first category, ML Lifecycles, includes Data Collection, Data Preparation, Feature Engineering, Model Training, Model Tuning, Model in Production, and Monitoring, each represented by a separate mini-graphic.

The second category, Key MLOps Challenges, lists Quality Control, Scalability, and Automation as the main challenges.

The last category, Most Used Tool Set, includes the following tools: Jupyter, GitHub, DataBricks, TensorFlow, PyTorch, Hadoop, Docker, Kubernetes, Jenkins, Airflow, and ELK stack. Each tool is represented by its respective logotype.
(.venv) bharathkumardasaraju@5.Vision$

Object Detection in Images

Detecting objects with Vision API

import os
from openai import OpenAI

# Initialize OpenAI client
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))  # Ensure API key is set


def detect_objects(image_url):
    """
    Uses OpenAI's GPT model to analyze an image and describe the detected objects.
    """
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": f"Detect objects in this image and describe them: {image_url}"}],
        max_tokens=300
    )

    return response.choices[0].message.content


# Example Image URL (Replace with an actual image URL)
image_url = "https://github.com/Bharathkumarraju/learn-ai-llm-ml-ops/blob/main/1.ml-ops-tools/tools-for-mlops.png"

# Run object detection
detected_objects = detect_objects(image_url)

# Print detected objects
print("Detected Objects:", detected_objects)

Image Question & Answer

Asking questions about images

import openai

def visual_question_answering(image_url, question):
    response = openai.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": f"Here is an image: {image_url}. Question: {question}"}],
        max_tokens=100
    )
    return response.choices[0].message.content

image_url = "https://github.com/Bharathkumarraju/learn-ai-llm-ml-ops/blob/main/1.ml-ops-tools/tools-for-mlops.png"
question = "What is this object?"
answer = visual_question_answering(image_url, question)
print("Answer:", answer)

Image Generation using DALL·E

Generate images with DALL·E

import openai

def generate_image_from_sketch(image_url, text_description):
    response = openai.images.generate(
        model="dall-e-3",
        prompt= f"Use the following image as a base: {image_url}. Add these details: {text_description}",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = "https://en.wikipedia.org/wiki/Sydney#/media/File:Sydney_Opera_House_and_Harbour_Bridge_Dusk_(2)_2019-06-21.jpg"
description = "Add a bright blue sky and detailed buildings in the background."
generated_image_url = generate_image_from_sketch(image_url, description)
print("Generated Image URL:", generated_image_url)

Advanced DALL·E Prompts

Another DALL·E example

from openai import OpenAI

client =  OpenAI()

def moderate_image(image_url):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content":f"Moderate this image and check if it is safe: {image_url}"}],
        max_tokens=100
    )
    return response.choices[0].message.content

image_url = "https://en.wikipedia.org/wiki/Brad_Pitt#/media/File:Brad_Pitt_Cannes_2012.jpg"
moderation_status = moderate_image(image_url)
print("Moderation Status:", moderation_status)

DALL·E Prompt Example 1

from openai import OpenAI
client = OpenAI()

def analyze_face(image_url):
    response = client.images.generate(
        model="dall-e-3",
        prompt= f"Analyze the face in this image and describe the person's age, gender, and emotion: {image_url}",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = "https://en.wikipedia.org/wiki/Lion#/media/File:Okonjima_Lioness.jpg"
face_analysis = analyze_face(image_url)
print("Face Analysis:", face_analysis)

DALL·E Prompt Example 2

from openai import OpenAI

client = OpenAI()

def image_to_image_translation(input_image_url, transformation_description):
    response = client.images.generate(
        model="dall-e-3",
        prompt= f"Transform the image at {input_image_url} by {transformation_description}",
        size = '1024x1024'
    )
    return response.data[0].url

input_image_url = "https://en.wikipedia.org/wiki/Lion#/media/File:Okonjima_Lioness.jpg"
transformation_description = "convert this sketch into a photorealistic image."
output_image_url = image_to_image_translation(input_image_url, transformation_description)
print("Translated Image URL:", output_image_url)

AI Image Samples

DALL·E Prompt Example 3

from openai import OpenAI

client = OpenAI()

def photo_gen():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "Create a cartoon style photo of a Singaporean person trying to become a YouTuber..Republic of singapore!",
        size = '1024x1024'
    )
    return response.data[0].url

print(photo_gen())

Practical DALL·E Applications

from openai import OpenAI

client = OpenAI()

def generate_ad_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "A futuristic running shoe with neon highlights on a misty mountain trail.",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = generate_ad_visual()
print("Generated Image URL:", image_url)

from openai import OpenAI
client = OpenAI()

def generate_insta_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "A luxury makeup palette displayed on a marble countertop with soft lighting.",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = generate_insta_visual()
print("Generated Image URL:", image_url)

from openai import OpenAI
client = OpenAI()

def generate_cert_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "AI certifications both in AWS Cloud and Micorsoft AZURE with colorful  neon highlights on a misty mountain trail as background with space ",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = generate_cert_visual()
print("Generated Image URL:", image_url)

from openai import OpenAI
client = OpenAI()


def generate_science_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "An anatomy of an animal cell with a white background.",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = generate_science_visual()
print("Generated Image URL:", image_url)

from openai import OpenAI
client = OpenAI()

def generate_game_char_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "A visual character of an animated game with a very cool outfit.",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = generate_game_char_visual()
print("Generated Image URL:", image_url)

from openai import OpenAI
client = OpenAI()

def generate_product_test_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "A luxury wooden table and a chair with some gadgets.",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = generate_product_test_visual()
print("Generated Image URL:", image_url)

from openai import OpenAI
client = OpenAI()

def generate_concept_home_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt="A luxury 3D architecture of a modern house with a glass facade and roof top garden",
        size='1024x1024'
    )
    return response.data[0].url

image_url = generate_concept_home_visual()
print("Generated Image URL:", image_url)

from openai import OpenAI
client = OpenAI()

def generate_heart():
    response = client.images.generate(
        model="dall-e-3",
        prompt="A 3D illustration of the human heart showing the flow of blood through the chambers.",
        size='1024x1024'
    )
    return response.data[0].url

image_url = generate_heart()
print("Generated Image URL:", image_url)

CLIP for Vision Understanding

CLIP-based visual semantic understanding

from openai import OpenAI

client = OpenAI()

# function that can generate the visuals
def generate_picture():
    response = client.images.generate(
        model="dall-e-3",
        prompt="Playing golf on Mars with aliens",
        size="1024x1024"
    )
    return response.data[0].url

image_url = generate_picture()

print("Generated Image URL:", image_url)

Image Captioning with GPT/Vision

Generate captions from image content

from openai import OpenAI

client = OpenAI()

# image url
image_url = "https://assets-prd.ignimgs.com/2022/06/10/netflix-one-piece-1654901410673.jpg"


# function that can generate captions
def generate_captions(image_url):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": [{"type": "text", "text": "What is this image?"},
                                               {"type": "image_url", "image_url": {"url": image_url}}]}],
        max_tokens=125
    )

    return response.choices[0].message.content


print(generate_captions(image_url))

Penguin Identification (Fun Demo)

Penguin classifier via OpenAI

from openai import OpenAI
client = OpenAI()

def generate_game_char_visual():
    response = client.images.generate(
        model="dall-e-3",
        prompt= "A Penguin on the computer, vaporwave.",
        size = '1024x1024'
    )
    return response.data[0].url

image_url = generate_game_char_visual()
print("Generated Image URL:", image_url)

Prompting the Image

Interacting with images using prompts

from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "What’s in this image?"},
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg",
                    },
                },
            ],
        }
    ],
    max_tokens=300,
)

print(response.choices[0].message.content)

Object Detection Example

Detect objects in uploaded images

from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Examine this medical image. Explain what the injury is."},
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://media02.stockfood.com/largepreviews/MzU50Dg1NTQx/11609211-Broken-arm-X-ray.jpg",
                    },
                },
            ],
        }
    ],
    max_tokens=300,
)

print(response.choices[0])

Vehicle Detection Sample

Classify and tag cars from image input

from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "You are an autonomous vehicle. What do you detect in front of you?"},
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://media.istockphoto.com/id/636690722/photo/driving-at-sunset-view-from-the-driver-angle-car-focusinside.jpg?s=612x612&w=0&k=20&c=B-D5L7GVi93AhhjfoLngbxHB8AEBjXPK_ZQ8tZEmSBo="
                    },
                },
            ],
        }
    ],
    max_tokens=300,
)

print(response.choices[0])

Vision Demos (Shell Scripts)

Vision API Shell Demo

Basic Vision API workflow in shell

# vision API

(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./1.openai-vision.py
Image Caption: This image is an infographic representing various tools for Machine Learning Operations (MLOps). The infographic is divided into three main categories: ML (Machine Learning) Lifecycles, Key MLOps Challenges, and Most Used Tool Set.

The first category, ML Lifecycles, includes Data Collection, Data Preparation, Feature Engineering, Model Training, Model Tuning, Model in Production, and Monitoring, each represented by a separate mini-graphic.

The second category, Key MLOps Challenges, lists Quality Control, Scalability, and Automation as the main challenges.

The last category, Most Used Tool Set, includes the following tools: Jupyter, GitHub, DataBricks, TensorFlow, PyTorch, Hadoop, Docker, Kubernetes, Jenkins, Airflow, and ELK stack. Each tool is represented by its respective logotype.
(.venv) bharathkumardasaraju@5.Vision$


Detect objects in a image




(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./4.openai-vision-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-wgDpReIt5T1SsZyxNH4Wo8AA.png?st=2025-03-15T02%3A11%3A28Z&se=2025-03-15T04%3A11%3A28Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A27%3A52Z&ske=2025-03-15T18%3A27%3A52Z&sks=b&skv=2024-08-04&sig=FF6jm47%2BpJhhBXLtEDRyCerJ5W7STdXFIqlRVRXfSZM%3D
(.venv) bharathkumardasaraju@5.Vision$



(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./7.openai-dalle.py
Translated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-o4zrg8sD9y5uXAOkRtabJXAo.png?st=2025-03-15T02%3A27%3A15Z&se=2025-03-15T04%3A27%3A15Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A33%3A14Z&ske=2025-03-15T18%3A33%3A14Z&sks=b&skv=2024-08-04&sig=Q3maWlX5BWBquL1om3CRHYAe4qxqAnClFULB4CcHkRA%3D
(.venv) bharathkumardasaraju@5.Vision$

Image Generation (DALL·E) Script

Generate image from prompts using shell

DALL-E is an AI model that generate images basedon input text.

(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./8.delle-generate-image.py
https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-BS54ehxwtd8LBMzdnyEpKjqj.png?st=2025-03-15T05%3A56%3A30Z&se=2025-03-15T07%3A56%3A30Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A12%3A38Z&ske=2025-03-15T18%3A12%3A38Z&sks=b&skv=2024-08-04&sig=dooBL%2B8jsCygX/gewO4FqTjZ%2BXe8wm4mMlfWWykjaaU%3D
(.venv) bharathkumardasaraju@5.Vision$

Practical Applications of DALL·E

Practical DALL·E integration examples

DALL-E - Practical Applications -- For Everyday usecases

(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./8.delle-generate-image.py
https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-BS54ehxwtd8LBMzdnyEpKjqj.png?st=2025-03-15T05%3A56%3A30Z&se=2025-03-15T07%3A56%3A30Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A12%3A38Z&ske=2025-03-15T18%3A12%3A38Z&sks=b&skv=2024-08-04&sig=dooBL%2B8jsCygX/gewO4FqTjZ%2BXe8wm4mMlfWWykjaaU%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./8.delle-generate-image.py
https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-QUcQQYTH0UToop0EM5SUGaJ5.png?st=2025-03-15T05%3A58%3A30Z&se=2025-03-15T07%3A58%3A30Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A17%3A39Z&ske=2025-03-15T18%3A17%3A39Z&sks=b&skv=2024-08-04&sig=NTKcE1D1XgH4pmzOpGnTZ9J0/tBQNiLTPks5iDID8Do%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./8.delle-generate-image.py
https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-52Hc6yJx8zzoi9ND2hLdDHPA.png?st=2025-03-15T05%3A59%3A22Z&se=2025-03-15T07%3A59%3A22Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A30%3A12Z&ske=2025-03-15T18%3A30%3A12Z&sks=b&skv=2024-08-04&sig=Kw%2BrrwlaK0guND15ias/wIMZrtNm2nVQzNVaLLMf2wI%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./9.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-6dI25zG7qRjbSu7PFQ6pfo3G.png?st=2025-03-15T06%3A13%3A39Z&se=2025-03-15T08%3A13%3A39Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A20%3A05Z&ske=2025-03-15T18%3A20%3A05Z&sks=b&skv=2024-08-04&sig=qwsfmZMiWPNot5W1v4w78j4HTpLYS8cYqwdLqaN089w%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./10.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-JNsPc0rZ41iDoI9z6G554yKD.png?st=2025-03-15T06%3A17%3A56Z&se=2025-03-15T08%3A17%3A56Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A20%3A59Z&ske=2025-03-15T18%3A20%3A59Z&sks=b&skv=2024-08-04&sig=qmoYRHoRlMmvlLQjTDxbblI2ZDuDUEAnL7c6qFpHF0k%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./11.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-TbvoAOCPLPKjJi2exKsVOYhV.png?st=2025-03-15T06%3A21%3A29Z&se=2025-03-15T08%3A21%3A29Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A12%3A43Z&ske=2025-03-15T18%3A12%3A43Z&sks=b&skv=2024-08-04&sig=axgYdngJAxWPimrtGMAydj4RJg13b2mIONyVlA/wkjM%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./11.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-8NX71IO18mShZRCwOV06nqq0.png?st=2025-03-15T06%3A22%3A56Z&se=2025-03-15T08%3A22%3A56Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A36%3A27Z&ske=2025-03-15T18%3A36%3A27Z&sks=b&skv=2024-08-04&sig=%2BSJ4IR3fOf4iViAPzj5rfDA6dMCGcvWfVMIPLDepiEA%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./12.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-m5bueGXvQ2TixK0LZFAqGxMi.png?st=2025-03-15T06%3A25%3A46Z&se=2025-03-15T08%3A25%3A46Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A18%3A36Z&ske=2025-03-15T18%3A18%3A36Z&sks=b&skv=2024-08-04&sig=Cr2FBT4NJIUWHLT71tnBH0DiRNuP%2BV0ic6AyGGvNkF8%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./13.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-Nn00LZgH2qyiGgqWH7bsGIr1.png?st=2025-03-15T06%3A28%3A39Z&se=2025-03-15T08%3A28%3A39Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A17%3A38Z&ske=2025-03-15T18%3A17%3A38Z&sks=b&skv=2024-08-04&sig=6At8QmcNh2SW2bctqqGz/0QaAFiBtUbUI3tMQK3MoOg%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./14.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-fPP4aic6VytXLUdNxVbuVpPc.png?st=2025-03-15T06%3A31%3A55Z&se=2025-03-15T08%3A31%3A55Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T22%3A33%3A30Z&ske=2025-03-15T22%3A33%3A30Z&sks=b&skv=2024-08-04&sig=9hMZQRmecNgNQ9mUkRgsMB0D4JBkLMycsj82Hl3QnUU%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./15.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-6nImocN8aMuwPakvWpRqdcGF.png?st=2025-03-15T06%3A34%3A25Z&se=2025-03-15T08%3A34%3A25Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A17%3A33Z&ske=2025-03-15T18%3A17%3A33Z&sks=b&skv=2024-08-04&sig=gLVLuMHWLPj4Lkws/UxwK5Z6EsZTnuCO69J5FqgWBEQ%3D
(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./16.practical-apps-dalle.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-vQAZBqnhskRICr3uVZ9g6Kkg.png?st=2025-03-15T06%3A41%3A40Z&se=2025-03-15T08%3A41%3A40Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T22%3A59%3A53Z&ske=2025-03-15T22%3A59%3A53Z&sks=b&skv=2024-08-04&sig=ENPkGXhLJOVxWElwFC0m9sLBcC0L31nmuIrIoEzfWeo%3D
(.venv) bharathkumardasaraju@5.Vision$

CLIP-Based Vision Script

Vision semantic search using CLIP

Understanding the relationship between Images and Text.

CLIP is a model to handle text patterns in the images.


(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./18.image_captions.py
This image is from the anime and manga series "One Piece." It features three characters from the series. The central character is wearing a straw hat, which is iconic to the series. "One Piece" follows the adventures of a group of pirates in search of a legendary treasure.
(.venv) bharathkumardasaraju@5.Vision$


(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./19.penguin.py
Generated Image URL: https://oaidalleapiprodscus.blob.core.windows.net/private/org-HX2Eo7f67VJnvWULLykeVcXx/user-D6Yk4QLHp5zfjnwikJee8IAi/img-ApYFmH9431P5UIlHpMAEBsdX.png?st=2025-03-15T07%3A40%3A00Z&se=2025-03-15T09%3A40%3A00Z&sp=r&sv=2024-08-04&sr=b&rscd=inline&rsct=image/png&skoid=d505667d-d6c1-4a0a-bac7-5c84a87759f8&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2025-03-14T18%3A18%3A41Z&ske=2025-03-15T18%3A18%3A41Z&sks=b&skv=2024-08-04&sig=3Wl48UO25Qylm3NiyHuxXKZx2vTB3qBxZJa1aXXq%2B3c%3D
(.venv) bharathkumardasaraju@5.Vision$



(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./20.prompting_the_image.py
Choice(finish_reason='stop', index=0, logprobs=None, message=ChatCompletionMessage(content='The image depicts a serene landscape featuring a wooden walkway or boardwalk that meanders through lush green grass and shrubs. In the background, there are trees and a partly cloudy blue sky, suggesting a peaceful outdoor setting, possibly in a marsh or nature reserve. The overall atmosphere is calm and picturesque, indicating a pleasant day.', refusal=None, role='assistant', audio=None, function_call=None, tool_calls=None, annotations=[]))
(.venv) bharathkumardasaraju@5.Vision$


(.venv) bharathkumardasaraju@5.Vision$ python3.12 ./20.prompting_the_image.py
The image depicts a serene landscape featuring a wooden pathway that winds through a verdant field. The grass appears lush and vibrant, with patches of greenery and some distant trees visible. The sky above is bright and blue, dotted with soft, white clouds. Overall, it creates a peaceful, natural scene.
(.venv) bharathkumardasaraju@5.Vision$