diff --git a/docs/source/reducing_memory_usage.md b/docs/source/reducing_memory_usage.md index e6f33f6ebd..ecc921168e 100644 --- a/docs/source/reducing_memory_usage.md +++ b/docs/source/reducing_memory_usage.md @@ -44,7 +44,7 @@ training_args = DPOConfig(..., max_completion_length=...) -SFT truncation is applied to the input sequence via the `max_length` parameter. +SFT truncation is applied to the input sequence via the `max_seq_length` parameter.
Truncation input ids @@ -92,4 +92,4 @@ training_args = SFTConfig(..., packing=True, max_seq_length=512) Packing may cause batch contamination, where adjacent sequences influence one another. This can be problematic for some applications. For more details, see [#1230](https://github.com/huggingface/trl/issues/1230). - \ No newline at end of file +