Some common quotes should you know about ChatGPT

1. text-davinci-003  the most capable model in chatGPT-3 series. It can process up to max 4000 tokens as per request.
2. The key concepts and techniques that are fundamental to using the API for any task, including:
   1. Content generation
   2. Summarization
   3. Classification, categorization, and sentiment analysis
   4. Data extraction
   5. Translation
   6. Many more!
      
      
3. You input some text as a prompt, and the API will return a text completion that attempts to match whatever instructions or context you gave it.
   1. Prompt: Write a tagline for an ice cream shop.
   2. Completion: We serve up smiles with every scoop!
4. This model can understand and process by breaking down words into tokens. Tokens can be words or chunks of character.For example, the word “hamburger” gets broken up into the tokens “ham”, “bur” and “ger”, while a short and common word like “pear” is a single token. Many tokens start with a whitespace, for example “ hello” and “ bye”.
5. one token generally corresponds to ~4 characters of text (so 100 tokens ~= 75 words).
6. generated completion combined must be no more than the model's maximum context length (for most models this is 2048 tokens, or about 1500 words).
7. GPT-3 models are called Davinci, Curie, Babbage and Ada.
8. Start with proper instruction and also try to give examples. suppose-
   1. Suggest one name for a horse. Ans: lighting
   2. Suggest one name for a black horse. Ans: Midnight
      
      
9. In many cases, it’s helpful to both show and tell the model what you want. Adding examples to your prompt can help communicate patterns or nuances. Try submitting this prompt which includes a couple examples. Suggest three names for an animal that is a superhero:
   1. 1. Super Stallion 2. Captain Colt 3. Mighty Mustang
      
      
10. Suggest three names for an animal that is a superhero:
    1. Animal: Cat
    2. Names: Captain Sharpclaw, Agent Fluffball, The Incredible Feline
    3. Animal: Dog
    4. Names: Ruff the Protector, Wonder Canine, Sir Barks-a-Lot
    5. Animal: Horse
    6. Names:
    7. OpenAI answer: Super Stallion, Mighty Mare, The Magnificent Equine

1. The model would always return identical or very similar completions. This is because your temperature was set to 0. If you set it to 1 then it will give you a random result.
2. If you submit this prompt 4 times with temperature set to 0, the model will always return “ animal” next because it has the highest probability. If you increase the temperature, it will take more risks and consider tokens with lower probabilities.  A moderate temperature of 0.6 should work well.

1. IF TEMPERATURE IS 0:
2. Horses are my favorite animal
3. Horses are my favorite animal
4. Horses are my favorite animal
5. Horses are my favorite animal

IF TEMPERATURE IS 1:

• Horses are my favorite animal
• Horses are my favorite animals
• Horses are my favorite !
• Horses are my favorite animal

1. Pricing is pay-as-you-go per 1,000 tokens, with $18 in free credit that can be used during your first 3 months.
2. Community libraries for swift : https://github.com/dylanshine/openai-kit Please note that OpenAI does not verify the correctness or security of these projects.
3. Older versions of our GPT-3 models are available as davinci, curie, babbage, and ada.
   1. Davinci good at: Complex intent, cause and effect, summarization for audience.
   2. Curie good at: Language translation, complex classification, text sentiment, summarization.
   3. Babbage good at: Moderate classification, semantic search classification.
   4. Ada good at: Parsing text, simple classification, address correction, keywords
   5. Note: Any task performed by a faster model like Ada can be performed by a more powerful model like Curie or Davinci.
4. You can use the GPT comparison tool (​​https://gpttools.com/comparisontool) that lets you run different models side-by-side to compare outputs, settings, and response times and then download the data into a .xls Excel spreadsheet.

Appendix: Notes

1. We used the suggested settings for Chat: Temperature = 0.9, response length = 150, Top P = 1
2. Actual swear words were replaced with 🤬
3. Polite has its own cultural biases and baggage. Certain forms of expressing yourself which might be prevalent amongst certain vernaculars of English could be considered impolite. Useful to think of it like we're steering the model to no longer be obliviously impolite through prompt engineering