Llama 3-8B

Training compute

7.2×10²³ FLOP

Parameters

8B

Published

Apr 18, 2024

Llama 3-8B is an AI model developed by Meta AI (United States), first published in April 2024. It works in the language domain, on tasks such as chat, language modeling/generation, code generation and question answering.

Training it took an estimated 7.2×10²³ FLOP of compute (estimation method: operation counting). The model has 8,000,000,000 parameters. It was trained on roughly 15T datapoints. Training ran on NVIDIA H100 SXM5 80GB.

Access: Open weights (restricted use). Its weights are openly available. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Meta AI
Country of organization: United States
Domain: Language
Task: Chat, Language modeling/generation, Code generation, Question answering
Training compute: 7.2×10²³ FLOP
Compute estimation method: Operation counting
Parameters: 8,000,000,000
Dataset size: 15T
Training hardware: NVIDIA H100 SXM5 80GB
Chip-hours: 1.3M
Model accessibility: Open weights (restricted use)
Open weights: Yes
Epoch confidence: Confident

Benchmark results

01

Epoch Capabilities Index

ECI Score

116.4