Variational (untied weights, MC) LSTM (Large)

Training compute

5.9×10¹⁵ FLOP

Parameters

66M

Published

Dec 16, 2015

Variational (untied weights, MC) LSTM (Large) is an AI model developed by University of Cambridge (United Kingdom), first published in December 2015. It works in the language domain, on tasks such as language modeling.

Training it took an estimated 5.9×10¹⁵ FLOP of compute (estimation method: operation counting). The model has 66,000,000 parameters. It was trained on roughly 929K datapoints.

Access: Unreleased. Its weights are not openly released. The reference paper has 1,838 citations. Epoch AI rates the confidence of this record as confident.

Full record

Organization: University of Cambridge
Country of organization: United Kingdom
Domain: Language
Task: Language modeling
Training compute: 5.9×10¹⁵ FLOP
Compute estimation method: Operation counting
Parameters: 66,000,000
Dataset size: 929K
Numerical format: FP32
Model accessibility: Unreleased
Open weights: No
Citations: 1,838
Epoch confidence: Confident