https://github.com/feifeibear/dpskv3mfu

Estimate MFU for DeepSeekV3
https://github.com/feifeibear/dpskv3mfu

Last synced: 3 months ago
JSON representation

Estimate MFU for DeepSeekV3

• Host: GitHub
• URL: https://github.com/feifeibear/dpskv3mfu
• Owner: feifeibear
• Created: 2025-01-05T01:07:49.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2025-01-05T01:42:05.000Z (5 months ago)
• Last Synced: 2025-01-05T02:19:48.590Z (5 months ago)
• Language: Python
• Size: 1000 Bytes
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# DPSKV3MFU

Estimate MFU for DeepSeek V3 model.

The MFU is around 39.0% for 4K context length pretraining phase.