Experimental browser for the Atmosphere
Anyone know what proportion of training data going into MidJourney, Dall-E, etc. is images vs. language?
Dec 13, 2024, 9:27 PM
{ "uri": "at://did:plc:lllawdfris2txwt4h6c5fmei/app.bsky.feed.post/3ld7pyy3irs2h", "cid": "bafyreie6fqdjv6rpynej3aovjndpqe66sciwdjldtotyqjhkchxlfx5spe", "value": { "text": "Anyone know what proportion of training data going into MidJourney, Dall-E, etc. is images vs. language?", "$type": "app.bsky.feed.post", "langs": [ "en" ], "createdAt": "2024-12-13T21:27:02.412Z" } }