Update readme

README.md

@@ -19,7 +19,7 @@ The image generator operates in a 56x80 RGB space, while the condition generator
 For ease of use, the condition generator accepts arbitrary subsets of the tags instead of requiring the user to specify every tag. As in a [previous exploration](https://huggingface.co/gustproof/1girl-EDM2-XS-test-1), the 1024D vectors are encodings created by a [pretrained tagger](https://huggingface.co/SmilingWolf/wd-eva02-large-tagger-v3). The condition supports a total of ~9k rating/general/character tags, although not all of them are well-understood.
 
 The training dataset is a 1.1M-image `1girl solo` subset of the [Danbooru2023 dataset](https://huggingface.co/datasets/nyanko7/danbooru2023), downscaled to 56x80.
-The image generator is trained on 57M images, and the condition generator is trained on 268 images. The combined training can be completed in ~1.5 days on a 4090 GPU.
+The image generator is trained for 57M images, and the condition generator is trained for 268M images. The combined training can be completed in ~1.5 days on a 4090 GPU.
 
 ## Qualitative Results