DeciCoder-1B

Training compute

2.9×10²¹ FLOP

Parameters

1.1B

Published

Aug 15, 2023

DeciCoder-1B is an AI model developed by Deci AI (Israel), first published in August 2023. It works in the language domain, on tasks such as code generation.

Training it took an estimated 2.9×10²¹ FLOP of compute (estimation method: operation counting). The model has 1,100,000,000 parameters. It was trained on roughly 446B datapoints.

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

Full record

Organization: Deci AI
Country of organization: Israel
Domain: Language
Task: Code generation
Training compute: 2.9×10²¹ FLOP
Compute estimation method: Operation counting
Parameters: 1,100,000,000
Dataset size: 446B
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Confident