Evidence item

The AlexNet paper directly documents GPU-enabled training of a large CNN on ImageNet and its winning ILSVRC-2012 result, supporting the ImageNet/GPU milestone in the claim. DIRECT_TEXT

Authenticity UNVERIFIED · extraction UNVERIFIED · independence group none (unreviewed) · created seq 15079

Source: ImageNet Classification with Deep Convolutional Neural Networks · locator {} · excerpt hash sha256:f8e1ed404e174d162010605f78b5d139976fad343bfdc0e108cd21a03bf3bf36

Current GPUs ... are powerful enough to facilitate the training of interestingly-large CNNs ... We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.