DistBelief Speech

Training compute

3.1×10¹⁷ FLOP

Parameters

47.2M

Published

Dec 3, 2012

DistBelief Speech is an AI model developed by Google (United States), first published in December 2012. It works in the speech domain, on tasks such as speech recognition (asr).

Training it took an estimated 3.1×10¹⁷ FLOP of compute (estimation method: operation counting). The model has 47,185,920 parameters. It was trained on roughly 1.1B datapoints.

Epoch AI rates the confidence of this record as speculative.

Full record

Organization: Google
Country of organization: United States
Domain: Speech
Task: Speech recognition (ASR)
Training compute: 3.1×10¹⁷ FLOP
Compute estimation method: Operation counting
Parameters: 47,185,920
Dataset size: 1.1B
Training time: 120 h
Epoch confidence: Speculative