Open yewang opened 3 months ago
hi, sorry for the late response. I've been very busy recently and didn't notice the issue, I will check the code later.
hi, we updated the repo, which should have resolved the error. By the way, when calling smoothllm defense, the arguments must specify perturbation type via smoothllm_perturbations=RandomSwap/RandomInsert/RandomPatch
, detailed in config/defense/smoothllm.yaml
.
I'm trying to reproduce your results from your paper, but I'm having some issues when trying to run the code.
When I try to run:
I get the error message:
The file
config/task/AlpacaEval.yaml
specifiesattack_log_file: data/AlpavaEval/AlpacaEval.jsonl
, but this jsonl file does not exist. Instead,data/AlpacaEval/AlpacaEval-125.jsonl
is provided in the repo. However, changing that configuration file seems to not fix it, since the code actually specifiesdata/AlpavaEval/AlpacaEval.jsonl
again at line 203 oftasks.py
. Changing it there seems to fix the issue, but the code still winds up crashing due to another issue:I also noticed some other issues:
config/task/InstructionFollow.yaml
specifies_target_: AlpacaEval
, which seems incorrect.README.md
seems to have a typo in specifyingdefense=${ATTACK}
and is missing the attack parameter. I guess this part should beattack=${ATTACK} defense=${DEFENSE}
, like in line 25.I was able to get things to start running with the incontext defense (with the following command), but then it immediately drops into the interactive python debugger, and crashes if I try to continue the script.