From b5fad8c675d4fa9ea0a96c3746af1b047374b47b Mon Sep 17 00:00:00 2001 From: Victor Hall Date: Tue, 28 Feb 2023 21:19:52 -0500 Subject: [PATCH] docs --- doc/NOTDREAMBOOTH.md | 2 +- doc/OPTIMIZER.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/doc/NOTDREAMBOOTH.md b/doc/NOTDREAMBOOTH.md index 7187cb7..8a89a9b 100644 --- a/doc/NOTDREAMBOOTH.md +++ b/doc/NOTDREAMBOOTH.md @@ -20,7 +20,7 @@ I instead propose you replace images generated out of SD itself with original "g "Ground truth" for the sake of this document means real images not generated by AI. It's very easy to get publicly available ML data sets to serve this purpose and replace genarated "regularization" images with real photos or otherwise. -Sources include [FFHQ](https://github.com/NVlabs/ffhq-dataset), [Coco](https://cocodataset.org/#home), and [Laion](https://huggingface.co/datasets/laion/laion2B-en-aesthetic/tree/main). There is a simple scraper to search Laion parquet files in the tools repo, and the Laion dataset was used by Compvis and [Stability.ai](https://github.com/Stability-AI/stablediffusion#news) themselves to train the base model. +Sources include [FFHQ](https://github.com/NVlabs/ffhq-dataset), [Coco](https://cocodataset.org/#home), and [Laion](https://huggingface.co/datasets/laion/laion2B-en-aesthetic/tree/main). There is a simple scraper to search Laion parquet files in the tools repo, and the Laion dataset was used by [Compvis](https://github.com/CompVis/stable-diffusion#weights) and [Stability.ai](https://github.com/Stability-AI/stablediffusion#news) themselves to train the base model. Using ground truth allows fine tuning to scale to potentially as many images at you want, including training dozens of characters at once using thousands of images, which many EveryDream community members have done. diff --git a/doc/OPTIMIZER.md b/doc/OPTIMIZER.md index c07f695..dc2b598 100644 --- a/doc/OPTIMIZER.md +++ b/doc/OPTIMIZER.md @@ -8,7 +8,7 @@ or in train.json "optimizer_config": "optimizer.json" -A default `optimizer.json` is supplied which you can modify +A default `optimizer.json` is supplied which you can modify. This has expanded tweaking. This doc is incomplete, but there is information on the web on betas and weight decay setting you can search for.