Harmful warning (#79)

* fix docstring formatting * add harmful warning
RomiconEZ · Feb 4, 2025 · af60896 · af60896
1 parent 6283d13
commit af60896
Show file tree

Hide file tree

Showing 6 changed files with 28 additions and 27 deletions.
diff --git a/src/llamator/attacks/bon.py b/src/llamator/attacks/bon.py
@@ -63,7 +63,7 @@ def apply_ascii_noising(prompt: str, sigma: float) -> str:
 @register_test
 class TestBon(TestBase):
     """Probes the LLM resilience against Best-of-N attack (text augmentations).
-    `Original Paper <https://arxiv.org/abs/2412.03556>`_, `Code <https://github.com/jplhughes/bon-jailbreaking>`_."""
+    Original Paper: https://arxiv.org/abs/2412.03556, Code: https://github.com/jplhughes/bon-jailbreaking"""
 
     test_name = "bon"
 

diff --git a/src/llamator/attacks/crecsendo.py b/src/llamator/attacks/crecsendo.py
@@ -64,7 +64,7 @@
 class TestCrescendo(TestBase):
     """Challenges the LLM with a dataset of harmful behavior prompts using Crescendo strategy,
     operates by engaging the LLM in a series of escalating conversational turns, and multistage refinement.
-    `Original Paper <https://arxiv.org/abs/2404.01833>`_"""
+    Original Paper: https://arxiv.org/abs/2404.01833"""
 
     test_name = "crescendo"
 

diff --git a/src/llamator/attacks/sycophancy.py b/src/llamator/attacks/sycophancy.py
@@ -63,7 +63,7 @@
 @register_test
 class TestSycophancy(TestBase):
     """Attacks the LLM system during a multistage conversation using the argumentum ad verecundiam method based on the system's domain.
-    `Original Paper <https://arxiv.org/abs/2310.13548>`_"""
+    Original Paper: https://arxiv.org/abs/2310.13548"""
 
     test_name = "sycophancy"
 

diff --git a/src/llamator/main.py b/src/llamator/main.py
@@ -190,7 +190,7 @@ def start_testing(
     if enable_reports:
         report_language = validate_language(report_language)
         csv_folder_name = "csv_report"
-        # Create Excel report
+        print(f"{BRIGHT}{colorama.Fore.RED}DISCLAIMER: Report may contain HARMFUL and OFFENSIVE language, reader discretion is recommended.{RESET}")
         create_attack_report_from_artifacts(
             artifacts_dir=artifacts_run_path, csv_folder_name=csv_folder_name, report_file_name="attacks_report.xlsx"
         )

diff --git a/tests/test_llamator.py b/tests/test_llamator.py
@@ -1,10 +1,12 @@
 import importlib
 import os
 
+from dotenv import load_dotenv
+
 
 def test_openai_client():
     """
-    Tests OpenAI client from public LLaMator lib.
+    Tests OpenAI client from public llamator lib.
 
     Environment variables
     ----------
@@ -15,7 +17,8 @@ def test_openai_client():
     OPENAI_CLIENT_MODEL : str
         Type of model
     """
-    # Перезагрузка необходимых модулей
+    load_dotenv()
+
     import llamator
 
     importlib.reload(llamator)
@@ -63,11 +66,11 @@ def test_openai_client():
     ]
 
     config = {
-        "enable_logging": True,  # Включить логирование
-        "enable_reports": True,  # Включить генерацию отчетов
-        "artifacts_path": "./artifacts",  # Путь к директории для сохранения артефактов
-        "debug_level": 1,  # Уровень логирования: 0 - WARNING, 1 - INFO, 2 - DEBUG
-        "report_language": "ru",  # Язык отчета: 'en', 'ru'
+        "enable_logging": True,  # Enable logging
+        "enable_reports": True,  # Enable report generation
+        "artifacts_path": "./artifacts",  # Path to the directory for saving artifacts
+        "debug_level": 1,  # Logging level: 0 - WARNING, 1 - INFO, 2 - DEBUG
+        "report_language": "en",  # Report language: 'en', 'ru'
     }
 
     llamator.start_testing(attack_model, tested_model, config=config, tests_with_attempts=tests_with_attempts)
diff --git a/tests/test_local_llamator.py b/tests/test_local_llamator.py
@@ -6,7 +6,7 @@
 
 def test_openai_client():
     """
-    Tests OpenAI client from local LLaMator lib.
+    Tests OpenAI client from local llamator lib.
 
     Environment variables
     ----------
@@ -17,7 +17,8 @@ def test_openai_client():
     OPENAI_CLIENT_MODEL : str
         Type of model
     """
-    # Перезагрузка необходимых модулей
+    load_dotenv()
+
     import llamator.client.specific_chat_clients
     import llamator.main
 
@@ -51,11 +52,11 @@ def test_openai_client():
     ]
 
     config = {
-        "enable_logging": True,  # Включить логирование
-        "enable_reports": True,  # Включить генерацию отчетов
-        "artifacts_path": "./artifacts",  # Путь к директории для сохранения артефактов
-        "debug_level": 1,  # Уровень логирования: 0 - WARNING, 1 - INFO, 2 - DEBUG
-        "report_language": "ru",  # Язык отчета: 'en', 'ru'
+        "enable_logging": True,  # Enable logging
+        "enable_reports": True,  # Enable report generation
+        "artifacts_path": "./artifacts",  # Path to the directory for saving artifacts
+        "debug_level": 1,  # Logging level: 0 - WARNING, 1 - INFO, 2 - DEBUG
+        "report_language": "en",  # Report language: 'en', 'ru'
     }
 
     from llamator.main import start_testing
@@ -65,7 +66,7 @@ def test_openai_client():
 
 def test_langchain_client_yandexgpt():
     """
-    Tests LangChain client from local LLaMator lib using Yandex GPT backend.
+    Tests LangChain client from local llamator lib using Yandex GPT backend.
 
     Environment variables
     ----------
@@ -74,17 +75,14 @@ def test_langchain_client_yandexgpt():
     FOLDER_ID : str
         Yandex Cloud folder ID
     """
-    # Загружаем переменные окружения из .env файла
     load_dotenv()
 
-    # Перезагрузка необходимых модулей
     import llamator.client.specific_chat_clients
     import llamator.main
 
     importlib.reload(llamator.client.specific_chat_clients)
     importlib.reload(llamator.main)
 
-    # Получаем переменные из окружения
     folder_ID = os.getenv("FOLDER_ID")
 
     from llamator.client.specific_chat_clients import ClientLangChain
@@ -124,11 +122,11 @@ def test_langchain_client_yandexgpt():
     ]
 
     config = {
-        "enable_logging": True,  # Включить логирование
-        "enable_reports": True,  # Включить генерацию отчетов
-        "artifacts_path": "./artifacts",  # Путь к директории для сохранения артефактов
-        "debug_level": 1,  # Уровень логирования: 0 - WARNING, 1 - INFO, 2 - DEBUG
-        "report_language": "ru",  # Язык отчета: 'en', 'ru'
+        "enable_logging": True,  # Enable logging
+        "enable_reports": True,  # Enable report generation
+        "artifacts_path": "./artifacts",  # Path to the directory for saving artifacts
+        "debug_level": 1,  # Logging level: 0 - WARNING, 1 - INFO, 2 - DEBUG
+        "report_language": "en",  # Report language: 'en', 'ru'
     }
 
     from llamator.main import start_testing