Code Llama-70B

Training compute

1.3×10²⁴ FLOP

Parameters

70B

Published

Jan 29, 2024

Code Llama-70B is an AI model developed by Meta AI (United States), first published in January 2024. It works in the language domain, on tasks such as code generation.

Training it took an estimated 1.3×10²⁴ FLOP of compute (estimation method: operation counting). The model has 70,000,000,000 parameters. It was trained on roughly 3T datapoints. Training ran on 400 NVIDIA A100 SXM4 80 GB for about 6.5K hours.

Access: Open weights (restricted use). Its weights are openly available. It is built on top of Llama 2-70B. The reference paper has 3,163 citations. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Meta AI
Country of organization: United States
Domain: Language
Task: Code generation
Training compute: 1.3×10²⁴ FLOP
Compute estimation method: Operation counting
Parameters: 70,000,000,000
Dataset size: 3T
Training hardware: NVIDIA A100 SXM4 80 GB
Chips used: 400
Training time: 6,480 h
Chip-hours: 2.7M
Training power draw: 316.8 kW
Model accessibility: Open weights (restricted use)
Open weights: Yes
Base model: Llama 2-70B
Citations: 3,163
Epoch confidence: Confident