fix: Add missing "Jailbreak" category from example

#6
by jonatanvm - opened
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -97,7 +97,7 @@ model = AutoModelForCausalLM.from_pretrained(
97
  )
98
  def extract_label_categories_refusal(content):
99
  safe_pattern = r"Safety: (Safe|Unsafe|Controversial)"
100
- category_pattern = r"(Violent|Non-violent Illegal Acts|Sexual Content or Sexual Acts|PII|Suicide & Self-Harm|Unethical Acts|Politically Sensitive Topics|Copyright Violation|None)"
101
  refusal_pattern = r"Refusal: (Yes|No)"
102
  safe_label_match = re.search(safe_pattern, content)
103
  refusal_label_match = re.search(refusal_pattern, content)
 
97
  )
98
  def extract_label_categories_refusal(content):
99
  safe_pattern = r"Safety: (Safe|Unsafe|Controversial)"
100
+ category_pattern = r"(Violent|Non-violent Illegal Acts|Sexual Content or Sexual Acts|PII|Suicide & Self-Harm|Unethical Acts|Politically Sensitive Topics|Copyright Violation|Jailbreak|None)"
101
  refusal_pattern = r"Refusal: (Yes|No)"
102
  safe_label_match = re.search(safe_pattern, content)
103
  refusal_label_match = re.search(refusal_pattern, content)