Open xu3kev opened 5 months ago
I met the same problem. Have you tried any good solutions since then?
I'm facing this issue now. It has been two months since this was reported. Did anybody find a solution?
Also running into this
Ran into this and setting load_in_8bit to false made it work.
same here.
Please check that this issue hasn't been reported before.
Expected Behavior
should be able to do training as usual
Current behaviour
crash with the following error message
Steps to reproduce
run the codellama-7b lora example with deepspeed zero3
Config yaml
Possible solution
No response
Which Operating Systems are you using?
Python Version
3.10
axolotl branch-commit
main/c2b64e4dcff59cfbd754626e5172688433cc13e1
Acknowledgements