More hints on what to do with CUDA Out of memory errors (#925)

This commit is contained in:
Juraj Bednar
2023-12-13 08:38:38 +01:00
committed by GitHub
parent 5f79b8242f
commit b0cf397ecb

View File

@@ -1022,6 +1022,10 @@ Please reduce any below
- `gradient_accumulation_steps`
- `sequence_len`
If it does not help, try running without deepspeed and without accelerate (replace "accelerate launch" with "python") in the command.
Using adamw_bnb_8bit might also save you some memory.
> `failed (exitcode: -9)`
Usually means your system has run out of system memory.