Cosmos-1.0-Prompt-Upsampler-12B-Text2World

Cosmos | Code | Paper

Description:

Cosmos-1.0-Prompt-Upsampler-12B-Text2World is a large language model (LLM) designed to transform original input prompts into more detailed and enriched versions. It improves the prompts by adding more details and maintaining a consistent description structure before they are used in a text-to-world model, which generally leads to higher quality outputs. This model is ready for commercial use.

License:

This model is released under the NVIDIA Open Model License. For a custom license, please contact cosmos-license@nvidia.com.

Under the NVIDIA Open Model License, NVIDIA confirms:

• Models are commercially usable.
• You are free to create and distribute Derivative Models.
• NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models.

Important Note: If you bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or associated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism contained in the Model, your rights under NVIDIA Open Model License Agreement will automatically terminate.

Model Architecture:

Architecture Type: Transformer
Network Architecture: Mistral-NeMo-12B-Instruct

Input:

Input Type(s): Text
Input Format: String
Input Parameters: One-dimensional (1D)
Other Properties Related to Input: Max of 512 tokens

Output:

Output Type(s): Text
Output Format: String
Output Parameters: One-dimensional (1D)
Other Properties Related to Output: Max of 512 tokens

Software Integration:

Runtime Engine(s):

• Cosmos

Supported Hardware Microarchitecture Compatibility:

• NVIDIA Ampere
  
• NVIDIA Hopper
  Supported Operating System(s): Linux
  

Model Version:

The initial release (v1.0) of Cosmos Prompt Upsampler contains the following model:

• Cosmos-1.0-Prompt-Upsampler-12B-Text2World

Usage

See Cosmos for on how to use the model.

Example:

• Input: "A dog is playing with a ball."
• Output: "The video is a close-up, dynamic scene featuring a dog interacting with a ball. The setting is a simple, indoor environment with a plain, light-colored wall in the background. The dog, a medium-sized breed with a light brown coat and a white chest, is the central focus. It is actively engaged in playing with a small, red ball. The dog's ears are perked up, and its mouth is open, indicating excitement and playfulness. Throughout the video, the dog is seen holding the ball in its mouth, occasionally shaking its head and moving its body in a playful manner. The camera remains static, capturing the dog's movements from a consistent angle, allowing the viewer to focus on the dog's expressions and actions. The lighting is bright and even, highlighting the dog's fur and the vibrant color of the ball. The dog's tail is wagging, adding to the sense of joy and energy in the scene. The video captures the essence of a playful interaction between a pet and its owner, emphasizing the bond and the simple joy of playtime."

Ethical Considerations

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.

For more detailed information on ethical considerations for this model, please see the subcards of Explainability, Bias, Safety & Security, and Privacy below. Please report security vulnerabilities or NVIDIA AI Concerns here.