AI models
StarCoder 2 3B
Training compute
5.9×10²² FLOP
Parameters
3B
Published
Feb 29, 2024
StarCoder 2 3B is an AI model developed by Hugging Face, ServiceNow, NVIDIA and BigCode (United States), first published in February 2024. It works in the language domain, on tasks such as code generation and code autocompletion.
Training it took an estimated 5.9×10²² FLOP of compute (estimation method: reported). The model has 3,000,000,000 parameters. It was trained on roughly 622.1B datapoints. Training ran on NVIDIA A100 SXM4 80 GB.
Access: Open weights (restricted use). Its weights are openly available. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- Hugging Face, ServiceNow, NVIDIA, BigCode
- Country of organization
- United States
- Domain
- Language
- Task
- Code generation, Code autocompletion
- Training compute
- 5.9×10²² FLOP
- Compute estimation method
- Reported
- Parameters
- 3,000,000,000
- Dataset size
- 622.1B
- Training hardware
- NVIDIA A100 SXM4 80 GB
- Model accessibility
- Open weights (restricted use)
- Open weights
- Yes
- Epoch confidence
- Confident
Benchmark results
| 01 | Epoch Capabilities Index | ECI Score | 88.4 |
More from Hugging Face,ServiceNow,NVIDIA,BigCode
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models