PrunedHub-GPT-OSS-20B-28x

Description

PrunedHub-GPT-OSS-20B-28x is an open-weight text-generation model with roughly 20B parameters published by GOBA-AI-Labs on the Hugging Face Hub. It has 307 downloads. Runs with llama-cpp.

Categories

• Productivity

Tags

• llama-cpp
• moe
• pruning
• expert-pruning
• mixture-of-experts
• goba-ai-labs
• prunedhub
• moe-stream

Capabilities

• Text generation
• Conversational
• GGUF quantised builds
• Library: llama-cpp
• 20B parameters