T

Story

OpenAI GPT-OSS models use MXFP4 to cut inference costs

• 7
• 0
• rntn
• 2 days ago
• theregister.com