9 votes

Stable Dreamfusion: An open source implementation of Google's text-to-3D synthesis

1 comment

  1. Jakobeha
    Link
    Still doesn't look very accurate based on the one example, and but nonetheless still a milestone and a step towards open-source consumer-grade text-to-3D

    Still doesn't look very accurate based on the one example, and

    Currently, 15000 training steps take about 5 hours to train on a V100 [very expensive GPU]

    but nonetheless still a milestone and a step towards open-source consumer-grade text-to-3D

    2 votes