From 16f9e28048bafd7e3cdf3b3b82d5d945d328641f Mon Sep 17 00:00:00 2001 From: PocketDocLabs Date: Fri, 9 Jun 2023 16:10:58 -0700 Subject: [PATCH] Update README.md to reflect current gradient checkpointing support Previously the readme stated gradient checkpointing was incompatible with 4-bit lora in the current implementation however this is no longer the case. I have replaced the warning with a link to the hugging face documentation on gradient checkpointing. --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 2787c7a13..2b14fe94b 100644 --- a/README.md +++ b/README.md @@ -387,7 +387,7 @@ train_on_inputs: false # don't use this, leads to wonky training (according to someone on the internet) group_by_length: false -# does not work with current implementation of 4-bit LoRA +# Whether to use gradient checkpointing https://huggingface.co/docs/transformers/v4.18.0/en/performance#gradient-checkpointing gradient_checkpointing: false # stop training after this many evaluation losses have increased in a row