Note

This model has a special pricing applied. Find out more

Introduction

Jurassic-2 (J2) series is AI21 Labs top-notch series of state-of-the-art Large Language Models. The Jurassic-2 family base consists of various LM sizes, including Large, Grande and Jumbo. Additionally, there are instruction-tuned LMs available in Grande and Jumbo sizes.

Jurassic-2 delivers exceptional performance, showcasing substantial enhancements in quality and introducing novel features such as instruction-following.

Jurassic-2 - Grande

With its advanced text generation capabilities, this model is specifically designed to excel in language tasks that involve higher levels of complexity. It offers a range of fine-tuning options that enable users to optimize the quality of outputs while still maintaining affordability and efficiency.

Run AI21 Jurassic2 Grande Complete with an API

Running the API with Clarifai's Python SDK

You can run the Jurassic2 Grande Complete Model API using Clarifai’s Python SDK.

Export your PAT as an environment variable. Then, import and initialize the API Client.

Find your PAT in your security settings.

export CLARIFAI_PAT={your personal access token}

import os

from clarifai.client.model import Model

input = '''what are the causes of pollution?'''

api_key = API_KEY
additional_command = "Generate a summary focusing on the next steps"
inference_params = dict(max_tokens= 512, top_k= 50,temperature= 0.3, api_key = api_key, top_p=1)

# Model Predict
model_prediction = Model("https://clarifai.com/ai21/complete/models/Jurassic2-Grande").predict_by_bytes(input.encode(), input_type="text", inference_params=inference_params)

print(model_prediction.outputs[0].data.text.raw)

You can also run a Jurassic2-Grande-Complete API using other Clarifai Client Libraries like Java, cURL, NodeJS, PHP, etc here.

Inference Parameters

• max_tokens: The maximum number of tokens the model will generate as part of the response. Note: Setting a low value may result in incomplete generations.
• temperature: A non-negative float that tunes the degree of randomness in generation. Lower temperatures mean less random generations.
• top_k: Ensures only the top k most likely tokens are considered for generation at each step.
• top_p: Ensures that only the most likely tokens, with total probability mass of p, are considered for generation at each step.

How to Use

By employing prompt engineering, J2 models can be utilized for practically any language-related undertaking. This involves creating a well-suited prompt that encompasses task descriptions and/or a handful of examples.

To facilitate the process of prompt development for optimal outcomes in your specific use-case, several illustrative examples from AIstudio are added for reference.

Marketing

Write a tweet to promote an event for Shmolt, a food and more delivery company.
##
Event: Christmas.
Tweet: Oven broke in the last minute and you have no turkey? No need to panic, with Shmolt delivery it will be at your door step by Christmas eve!
##
Event: Valentine's Day.
Tweet: Order her roses using Shmolt, so you can have time to stop and smell the roses #Valentines
##
Event: Summer Sale.
Tweet: In this heat, why get out of the AC when you have 10% off of all Shmolt deliveries?
##
Event: Mother's Day.
Tweet:

Classify Topics

Classify the following news article into one of the following topics:
1. World
2. Sports
3. Business
4. Science and Technology
Title:
Astronomers Observe Collision of Galaxies, Formation of Larger
Summary:
An international team of astronomers has obtained the clearest images yet of the merger of two distant clusters of galaxies, calling it one of the most powerful cosmic events ever witnessed.
The topic of this article is:
Science and Technology

##

Classify the following news article into one of the following topics:
1. World
2. Sports
3. Business
4. Science and Technology
Title:
Bomb Explodes Near U.S. Military Convoy (AP)
Summary:
AP - A car bomb exploded early Sunday near a U.S. military convoy on the road leading to Baghdad's airport, Iraqi police said, and a witness said two Humvees were destroyed.
The topic of this article is:
World

##

Classify the following news article into one of the following topics:
1. World
2. Sports
3. Business
4. Science and Technology
Title:
Maradona goes to Cuba
Summary:
The former Argentine football star, Diego Armando Maradona, traveled on Monday to Cuba to continue his treatment against his addiction to drugs.
The topic of this article is:
Sports

##

Classify the following news article into one of the following topics:
1. World
2. Sports
3. Business
4. Science and Technology
Title:
Duke earnings jump in third quarter
Summary:
Duke Energy Corp. reports third-quarter net income of  $389 million, or 41 cents per diluted share, sharply above earnings of  $49 million, or 5 cents per diluted share, in the same period last year.
The topic of this article is:
Business

##

Classify the following news article into one of the following topics:
1. World
2. Sports
3. Business
4. Science and Technology
Title:
D.C. Unveils Stadium Plan
Summary:
Rumors spread that Major League Baseball is edging closer to moving the Expos to Washington as D.C. officials announce plans for a stadium on the Anacostia waterfront.
The topic of this article is:

Simplify Jargon

The following sentences contain business jargon. Rewrite them using simple words.

Jargon: The fund managers hope to increase yields by taking on leverage.
Simple: The fund managers hope to get more return on their investments by borrowing money.##

Jargon: I need to finish due diligence on this company before I can decide.
Simple: I need to finish background research on this company before I can decide.##

Jargon: Can you please get this document over the wall?
Simple: Can you please send this document?##

Jargon: Their legal team would like us to open our kimono regarding last year's deals.
Simple:

Note

stopSequence Parameter of LLM is set to ##. Generation is terminated as soon as the model generates ##

Limitations

Input prompts + generation should not exceed 8191 tokens

Disclaimer

Please be advised that this model utilizes wrapped Artificial Intelligence (AI) provided by AI21 Labs (the "Vendor"). These AI models may collect, process, and store data as part of their operations. By using our website and accessing these AI models, you hereby consent to the data practices of the Vendor. We do not have control over the data collection, processing, and storage practices of the Vendor. Therefore, we cannot be held responsible or liable for any data handling practices, data loss, or breaches that may occur. It is your responsibility to review the privacy policies and terms of service of the Vendor to understand their data practices. You can access the Vendor's privacy policy and terms of service at https://www.ai21.com/privacy-policy.

We disclaim all liability with respect to the actions or omissions of the Vendor, and we encourage you to exercise caution and to ensure that you are comfortable with these practices before utilizing the AI models hosted on our site.