Stable Code 3B

Training compute

2.1×10²² FLOP

Parameters

2.8B

Published

Jan 9, 2024

Stable Code 3B is an AI model developed by Stability AI (United Kingdom), first published in January 2024. It works in the language domain, on tasks such as language modeling/generation and code generation.

Training it took an estimated 2.1×10²² FLOP of compute (estimation method: operation counting). The model has 2,796,431,360 parameters. Training ran on 256 NVIDIA A100 SXM4 40 GB.

Access: Open weights (non-commercial). Its weights are openly available. Epoch AI rates the confidence of this record as likely.

Full record

Organization: Stability AI
Country of organization: United Kingdom
Domain: Language
Task: Language modeling/generation, Code generation
Training compute: 2.1×10²² FLOP
Compute estimation method: Operation counting
Parameters: 2,796,431,360
Training hardware: NVIDIA A100 SXM4 40 GB
Chips used: 256
Training power draw: 202.9 kW
Model accessibility: Open weights (non-commercial)
Open weights: Yes
Epoch confidence: Likely