Getting Began with Google’s Palm API Utilizing Python

Introduction

As AI is taking up the world, Massive language fashions are in enormous demand in expertise. Massive Language Fashions generate textual content in a method a human does. They can be utilized to develop pure language processing (NLP) functions various from chatbots and textual content summarizers to translation apps, digital assistants, and so forth.

Google launched its next-generation mannequin named Palm 2. This mannequin excels in superior scientific and mathematical operations and is utilized in reasoning and language translations. This mannequin is skilled over 100+ spoken phrase languages and 20+ programming languages.

As it’s skilled in numerous programming languages, it may be used to translate one programming language to a different. For instance, if you wish to translate Python code to R or JavaScript code to TypeScript, and so forth., you’ll be able to simply use Palm 2 to do it for you. Aside from these, it might probably generate idioms and phrases and simply cut up a fancy activity into less complicated duties, making it a lot better than the earlier giant language fashions.

Studying Aims

• Introduction to Google’s Palm API
• Discover ways to entry Palm API by producing an API key
• Utilizing Python, learn to generate easy textual content utilizing textual content mannequin
• Discover ways to create a easy chatbot utilizing Python
• Lastly, we focus on how you can use Langchain with Palm API.

This text was printed as part of the Information Science Blogathon.

What’s Palm API?

Utilizing the Palm API, you’ll be able to entry the capabilities of Google’s Generative AI fashions and develop fascinating AI-powered functions. Nevertheless, if you wish to work together straight with the Palm 2 mannequin from the browser, you should utilize the browser-based IDE “MakerSuite”. However you’ll be able to entry the Palm 2 mannequin utilizing the Palm API to combine giant language fashions into your functions and construct AI-driven functions utilizing your organization’s knowledge.

Three totally different immediate interfaces are designed, and you will get began with anyone amongst them utilizing the Palm API. They’re:

• Textual content Prompts: You need to use the mannequin named “text-bison-001 to generate easy textual content. Utilizing textual content prompts, you’ll be able to generate textual content, generate code, edit textual content, retrieve info, extract knowledge, and so forth..
• Information Prompts: These let you assemble prompts in a tabular format.
• Chat Prompts: Chat prompts are used to construct conversations. You need to use the mannequin named “chat-bison-001” to make use of chat companies.

Find out how to Entry Palm API?

Navigate to the web site https://builders.generativeai.google/ and be part of the maker suite. You may be added to the waitlist and shall be given entry in all probability inside 24 hours.

Generate an API key:

1. You’ll want to get your individual API key to make use of the API. 
2. You’ll be able to join your utility to the Palm API and entry its companies utilizing the API key. 
3. As soon as your account is registered, it is possible for you to to generate it.
4. Subsequent, go forward and generate your API key, as proven within the screenshot beneath:

Save the API key as we’ll use it additional.

Setting the Surroundings

To make use of the API with Python, set up it utilizing the command:

pip set up google-generativeai

Subsequent, we configure it utilizing the API key that we generated earlier.

import google.generativeai as palm
palm.configure(api_key=API_KEY)

To listing the out there fashions, we write the beneath code:

fashions = [model for model in palm.list_models()]

for mannequin in fashions:
  print(mannequin.identify)

Output:

fashions/chat-bison-001

fashions/text-bison-001

fashions/embedding-gecko-001

Find out how to Generate Textual content?

We use the mannequin “text-bison-001” to generate textual content and move GenerateTextRequest. The generate_text() operate takes in two parameters i.e., a mannequin and a immediate. We move the mannequin as “text-bison-001,” and the immediate comprises the enter string.

Clarification:

• Within the instance beneath, we move the model_id variable with the mannequin identify and a immediate variable containing the enter textual content.
• We then move the model_id as mannequin and the immediate as immediate to the generate_text() methodology.
• The temperature parameter signifies how random the response shall be. In different phrases, if you need the mannequin to be extra inventive, you can provide it a worth of 0.3.
• Lastly, the parameter “max_tokens” signifies the utmost variety of tokens the mannequin’s output can include. A token can include roughly 4 tokens. Nevertheless, should you don’t specify, a default worth of 64 shall be assigned to it.

Instance 1

model_id="fashions/text-bison-001"
immediate=""'write a canopy letter for a knowledge science job applicaton. 
Summarize it to 2 paragraphs of fifty phrases every. '''

completion=palm.generate_text(
    mannequin=model_id,
    immediate=immediate,
    temperature=0.99,
    max_output_tokens=800,
)

print(completion.end result)

Output:

We outline some time loop that asks for enter and generates a reply. The response.final assertion prints the response.

model_id="fashions/chat-bison-001"

immediate="I need assistance with a job interview for a knowledge analyst job. Are you able to assist me?"

examples=[
    ('Hello', 'Hi there mr. How can I be assistant?'),
    ('I want to get a High paying Job','I can work harder')
]

response=palm.chat(messages=immediate, temperature=0.2, context="Communicate like a CEO", examples=examples)
for messages in response.messages:
  print(messages['author'],messages['content'])


whereas True:
  s=enter()
  response=response.reply(s)
  print(response.final)

Output:

Utilizing Palm API with LangChain

LangChain is an open-source framework that permits you to join giant language fashions to your functions. To make use of Palm API with langchain, we import GooglePalmEmbeddings from langchain.embeddings. The langchain has an embedding class that gives an ordinary interface for numerous textual content embedding fashions equivalent to OpenAI, HuggingFace, and so forth.

We move the prompts as an array, as proven within the beneath instance. Then, we name llm._generate() operate and move the prompts array as a parameter.

from langchain.embeddings import GooglePalmEmbeddings
from langchain.llms import GooglePalm

llm=GooglePalm(google_api_key=API_KEY)
llm.temperature=0.2

prompts=["How to Calculate the area of a triangle?","How many sides are there for a polygon?"]
llm_result= llm._generate(prompts)

res=llm_result.generations
print(res[0][0].textual content)
print(res[1][0].textual content)

Output:

Immediate 1

1.

**Discover the bottom and top of the triangle.

** The bottom is the size of the facet of the triangle that's parallel to the bottom, and the peak is the size of the road section that's perpendicular to the bottom and intersects the alternative vertex.

2.

**Multiply the bottom and top and divide by 2.

** The system for the realm of a triangle is A = 1/2 * b * h.

For instance, if a triangle has a base of 5 cm and a top of 4 cm, its space could be 1/2 * 5 * 4 = 10 cm2.

Immediate 2

3

Conclusion

On this article, we’ve got launched Google’s newest Palm 2 mannequin and the way it’s higher than the earlier fashions. We then realized how you can use Palm API with Python Programming Language. We then mentioned how you can develop easy functions and generate textual content and chats. Lastly, we lined how you can embed it utilizing Langchain framework.

Key Takeaways

• Palm API permits customers to develop functions utilizing Massive Language Fashions
• Palm API gives a number of text-generation companies, equivalent to textual content service to generate textual content and a chat service to generate chat conversations.
• Google generative-ai is the Palm API Python Library and will be simply put in utilizing the pip command.

Regularly Requested Questions

The media proven on this article is just not owned by Analytics Vidhya and is used on the Creator’s discretion.

Associated