Mirrors
/
EveryDream2trainer
mirror of https://github.com/victorchall/EveryDream2trainer.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

readme updates

This commit is contained in:
Victor Hall 2023-02-18 14:49:30 -05:00
parent b526cb5093
commit 025fb2557f
2 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
README.md
View File

@ -43,6 +43,8 @@ Make sure to check out the [tools repo](https://github.com/victorchall/EveryDrea
[Validation](doc/VALIDATION.md) - Use a validation split on your data to see when you are overfitting and tune hyperparameters
[Troubleshooting](doc/TROUBLESHOOTING.md)
## Cloud
[Free tier Google Colab notebook](https://colab.research.google.com/github/victorchall/EveryDream2trainer/blob/main/Train_Colab.ipynb)

2
doc/LOWVRAM.md
View File

@ -20,6 +20,6 @@ Keeping the batch_size low reduces VRAM use. This is a more "fine dial" on VRAM
The third is gradient accumulation, which does not reduce VRAM, but gives you a "virtual batch size multiplier" when you are not able to increase the batch_size directly.
--grad_accum 2
--grad_accum 4
This will combine the loss from multiple batches before applying updates. There is some small VRAM overhead to this but not as much as increasing the batch size. Increasing it beyond 2 does not continue to increase VRAM, only going from 1 to 2 seems to affect VRAM use, and by a small amount.