llama-3.1-8b-instruct-4x1-moe-switch-top1

Description

llama-3.1-8b-instruct-4x1-moe-switch-top1 is an open-weight text-generation model with roughly 8B parameters published by Jongbin-kr on the Hugging Face Hub. It has 415 downloads. Runs with transformers.

Community Metrics

Categories

• Coding
• Productivity

Tags

• upcycled_llama
• text-generation
• mixture-of-experts
• sparse-upcycling
• switch-transformer
• conversational
• custom_code

Capabilities

• Text generation
• Conversational
• Library: transformers
• 8B parameters

Information