))
Large data sources with bad data. Smaller data sources with good data. Which is better?
Manage episode 372324125 series 3442262
Innehåll tillhandahållet av Sam Holcman. Allt poddinnehåll inklusive avsnitt, grafik och podcastbeskrivningar laddas upp och tillhandahålls direkt av Sam Holcman eller deras podcastplattformspartner. Om du tror att någon använder ditt upphovsrättsskyddade verk utan din tillåtelse kan du följa processen som beskrivs här https://sv.player.fm/legal.
Large bad data. Obviously not good and some have recognized this as “model collapse” – the bad data causes more bad data to be generated. Small bad data – nothing needs to be said here. Small (vetted, provenance-known) data – perhaps within your enterprise walls – perhaps the way to go, until large, good data that is used for training appears. When will this happen? I do not think this is on the horizon.
155 episoder
Dela
