Gorilla-7b-hf-delta-v1: API Invocation LLM

Gorilla-7b-hf-delta-v1 is a 7 billion parameter language model, fine-tuned from LLaMA, designed to enable Large Language Models (LLMs) to effectively use tools by invoking APIs. Developed by Gorilla LLM (UC Berkeley), this model excels at translating natural language queries into semantically and syntactically correct API calls, significantly reducing hallucination in the process. It has been specifically trained to reliably use over 1,600 Hugging Face APIs.

Key Capabilities

• API Generation: Translates natural language requests into executable API code snippets.
• Reduced Hallucination: Demonstrates improved accuracy in generating valid API calls compared to general-purpose LLMs.
• Hugging Face API Focus: Optimized for invoking a wide range of Hugging Face APIs for tasks such as translation, image processing, and more.
• Direct Code Output: Generates code that can be directly integrated into workflows, as shown in examples for transformers pipeline usage.

Good For

• Developers looking to integrate LLMs with external tools and services via APIs.
• Automating tasks that require interaction with Hugging Face models and libraries.
• Building applications where an LLM needs to dynamically generate and execute code for specific functions.
• Research into tool-use and API invocation capabilities of large language models.