From b0dfbe70864d95099e53b9bf131ee2112b2ec8f5 Mon Sep 17 00:00:00 2001 From: Kohya S Date: Mon, 26 Jun 2023 21:20:49 +0900 Subject: [PATCH] update readme --- README.md | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index 265c9af9..df059120 100644 --- a/README.md +++ b/README.md @@ -30,6 +30,8 @@ Summary of the feature: - `--cache_text_encoder_outputs`: Cache the outputs of the text encoders. This option is useful to reduce the GPU memory usage. This option cannot be used with options for shuffling or dropping the captions. - `--no_half_vae`: Disable the half-precision (mixed-precision) VAE. VAE for SDXL seems to produce NaNs in some cases. This option is useful to avoid the NaNs. +`requirements.txt` is updated to support SDXL training. + ### Tips for SDXL training - The default resolution of SDXL is 1024x1024. @@ -40,6 +42,7 @@ Summary of the feature: - Use Adafactor optimizer. RMSprop 8bit or Adagrad 8bit may work. AdamW 8bit doesn't seem to work. - The LoRA training can be done with 12GB GPU memory. - `--train_unet_only` option is highly recommended for SDXL LoRA. Because SDXL has two text encoders, the result of the training will be unexpected. +- PyTorch 2 seems to use slightly less GPU memory than PyTorch 1. Example of the optimizer settings for Adafactor with the fixed learning rate: ``` @@ -54,7 +57,7 @@ learning_rate = 4e-7 # SDXL original learning rate These files do not contain requirements for PyTorch. Because the versions of them depend on your environment. Please install PyTorch at first (see installation guide below.) -The scripts are tested with PyTorch 1.12.1 and 1.13.0, Diffusers 0.10.2. +The scripts are tested with PyTorch 1.12.1 and 2.0.1, Diffusers 0.17.1. ## Links to how-to-use documents