List your AI tool on AI Orion for free. Submit a tool

StableBeluga2

Stable Beluga 2 is a Lao-like language model based on Llama2 70B, finetuned on an Orca-style dataset for English text generation.

About StableBeluga2

Stable Beluga 2 is a Llama2 70B model finetuned on an Orca style Dataset. The model is developed by Stability AI and is available to run through Hugging Face. The model is a reference for researchers and developers exploring chat-like language models and their capabilities in English.

Stable Beluga 2 provides usage guidance including a Python code snippet to start chatting with the model. It shows how to load the tokenizer and model, set a system prompt, and generate text with specific generation parameters. The page also indicates a recommended prompt format and example interaction text to illustrate how to interact with the model.

It lists model details including the model type and language, and notes the license information for the fine-tuned checkpoints.

The page discusses ethical considerations and limitations, acknowledging that the model carries risks, that results can be inaccurate or biased, and that testing should be tailored to specific applications before deployment.

Use cases

  • Researchers evaluating large language models for English text generation and instruction following.

Key features

  • Model type and base architecture
  • Training data approach
  • Language support
  • Usage guidance
  • Deployment note
  • License information

Pros and cons

Cons

Not deployed by any Inference Provider

Reviews

Loading reviews…

Frequently asked questions

What is Stable Beluga 2?
Stable Beluga 2 is a Llama2 70B model finetuned on an Orca style Dataset
How is data handled or licensed?
Fine-tuned checkpoints (Stable Beluga 2) is licensed under the STABLE BELUGA NON-COMMERCIAL COMMUNITY LICENSE AGREEMENT

Alternatives to StableBeluga2

More in AI Models