FTW (For The Win)

Training compute

3.5×10¹⁹ FLOP

Parameters

126M

Published

Jul 3, 2018

FTW (For The Win) is an AI model developed by DeepMind (United Kingdom), first published in July 2018. It works in the games domain, on tasks such as capture the flag.

Training it took an estimated 3.5×10¹⁹ FLOP of compute (estimation method: third-party estimation). The model has 126,001,330 parameters. It was trained on roughly 2B datapoints.

Access: Unreleased. Its weights are not openly released. The reference paper has 802 citations. Epoch AI rates the confidence of this record as speculative.

Full record

Organization: DeepMind
Country of organization: United Kingdom
Domain: Games
Task: Capture the flag
Training compute: 3.5×10¹⁹ FLOP
Compute estimation method: Third-party estimation
Parameters: 126,001,330
Dataset size: 2B
Model accessibility: Unreleased
Open weights: No
Citations: 802
Epoch confidence: Speculative