EveryDream2trainer/doc/TRAINING.md

Starting a training session

Here are some example commands to get you started, you can copy paste them into your command line and press enter. Make sure the last line does not have ^ but all other lines do.

First, open a command line, then make sure to activate the environment:

activate_venv.bat

You should see your command line show (venv) at the beginning of the line. If you don't, something went wrong with setup.

Running from a json config file

You can edit the example train.json file to your liking, then run the following command:

python train.py --config train.json

Be careful with editing the json file, as any syntax errors will cause the program to crash. You might want to use a json validator to check your file before running it. You can use an online validator such as https://jsonlint.com/ or look at it in VS Code.

One particular note is if your path to data_root or resume_ckpt has backslashes they need to use double \\ or single /. There is an example train.json in the repo root.

Running from the command line with arguments

I recommend you copy one of the examples below and keep it in a text file for future reference. Your settings are logged in the logs folder, but you'll need to make a command to start training.

Training examples:

Resuming from a checkpoint, 50 epochs, 6 batch size, 3e-6 learning rate, cosine scheduler, generate samples evern 200 steps, 10 minute checkpoint interval, adam8bit, and using the default "input" folder for training data:

python train.py --resume_ckpt "sd_v1-5_vae" ^
--max_epochs 50 ^
--data_root "input" ^
--lr_scheduler cosine ^
--project_name myproj ^
--batch_size 6 ^
--sample_steps 200 ^
--lr 3e-6 ^
--ckpt_every_n_minutes 10 ^
--useadam8bit

Training from SD2 512 base model, 18 epochs, 4 batch size, 1.2e-6 learning rate, constant LR, generate samples evern 100 steps, 30 minute checkpoint interval, adam8bit, using imagesin the x:\mydata folder, training at resolution class of 640:

python train.py --resume_ckpt "512-base-ema" ^
--data_root "x:\mydata" ^
--max_epochs 18 ^
--lr_scheduler constant ^
--project_name myproj ^
--batch_size 4 ^
--sample_steps 100 ^
--lr 1.2e-6 ^
--resolution 640 ^
--clip_grad_norm 1 ^
--ckpt_every_n_minutes 30 ^
--useadam8bit

Training from the "SD21" model on the "jets" dataset on another drive, for 50 epochs, 6 batch size, 1.5e-6 learning rate, cosine scheduler that will decay in 1500 steps, generate samples evern 100 steps, save a checkpoint every 20 epochs, and use AdamW 8bit optimizer:

python train.py --resume_ckpt "SD21" ^
--data_root "R:\everydream-trainer\training_samples\mega\gt\objects\jets" ^
--max_epochs 25 ^
--lr_scheduler cosine ^
--lr_decay_steps 1500 ^
--lr_warmup_steps 20 ^
--project_name myproj ^
--batch_size 6 ^
--sample_steps 100 ^
--lr 1.5e-6 ^
--save_every_n_epochs 20 ^
--useadam8bit 

Copy paste the above to your command line and press enter. Make sure the last line does not have ^ but all other lines do. If you want you can put the command all on one line and not use the ^ carats instead.

How to resume

Point your resume_ckpt to the path in logs like so:

--resume_ckpt "R:\everydream2trainer\logs\myproj20221213-161620\ckpts\myproj-ep22-gs01099" ^

Or use relative pathing:

--resume_ckpt "logs\myproj20221213-161620\ckpts\myproj-ep22-gs01099" ^

You should point to the folder in the logs per above if you want to resume rather than running a conversion back on a 2.0GB or 2.5GB pruned file if possible.