FIX: integration tests and ImageTarget refactor (#1293)

Author: rlundeen2

.env_example

@@ -41,8 +41,9 @@ AZURE_FOUNDRY_DEEPSEEK_KEY="xxxxx"
 AZURE_FOUNDRY_PHI4_ENDPOINT="https://xxxxx.models.ai.azure.com"
 AZURE_CHAT_PHI4_KEY="xxxxx"
 
-AZURE_FOUNDRY_MINSTRAL3B_ENDPOINT="https://xxxxx.eastus2.models.ai.azure.com"
-AZURE_CHAT_MINSTRAL3B_KEY="xxxxx"
+AZURE_FOUNDRY_MISTRAL_LARGE_ENDPOINT="https://xxxxx.services.ai.azure.com/openai/v1/"
+AZURE_FOUNDRY_MISTRAL_LARGE_KEY="xxxxx"
+AZURE_FOUNDRY_MISTRAL_LARGE_MODEL="Mistral-Large-3"
 
 GROQ_ENDPOINT="https://api.groq.com/openai/v1"
 GROQ_KEY="gsk_xxxxxxxx"

doc/code/targets/5_multi_modal_targets.ipynb

@@ -5,7 +5,11 @@
 #       extension: .py
 #       format_name: percent
 #       format_version: '1.3'
-#       jupytext_version: 1.17.3
+#       jupytext_version: 1.18.1
+#   kernelspec:
+#     display_name: pyrit (3.13.5)
+#     language: python
+#     name: python3
 # ---
 
 # %% [markdown]
@@ -35,14 +39,6 @@
 objective = "Give me an image of a raccoon pirate as a Spanish baker in Spain"
 
 img_prompt_target = OpenAIImageTarget()
-# For Azure OpenAI with Entra ID authentication enabled, use the following command instead. Make sure to run `az login` first.
-# from pyrit.auth import get_azure_openai_auth
-# endpoint = "https://your-endpoint.openai.azure.com"
-# target = OpenAIChatTarget(
-#     endpoint=endpoint,
-#     api_key=get_azure_openai_auth(endpoint),
-#     model_name="your-deployment-name"
-# )
 
 image_scorer = SelfAskTrueFalseScorer(
     chat_target=OpenAIChatTarget(),
@@ -98,57 +94,60 @@
 #
 # This example demonstrates how to use the video target to create a video from a text-based prompt.
 #
-# from pyrit.executor.attack import (
-#     AttackExecutor,
-#     AttackScoringConfig,
-#     ConsoleAttackResultPrinter,
-#     PromptSendingAttack,
-# )
-# from pyrit.prompt_target import OpenAIChatTarget, OpenAIVideoTarget
-# from pyrit.score import (
-#     AzureContentFilterScorer,
-#     SelfAskTrueFalseScorer,
-#     TrueFalseQuestion,
-#     VideoFloatScaleScorer,
-#     VideoTrueFalseScorer,
-# )
-# from pyrit.setup import IN_MEMORY, initialize_pyrit_async
-#
-# await initialize_pyrit_async(memory_db_type=IN_MEMORY)  # type: ignore
-#
-# objectives = [
-#     "Video of a raccoon pirate eating flan at a cafe in Spain",
-#     "Video of a raccoon pirate eating a croissant at a cafe in France",
-# ]
-#
-# video_target = OpenAIVideoTarget()
-# objective_scorer = SelfAskTrueFalseScorer(
-#     chat_target=OpenAIChatTarget(),
-#     true_false_question=TrueFalseQuestion(true_description="A raccoon dressed as a pirate is actively eating a pastry"),
-# )
-#
-# video_scorer = VideoTrueFalseScorer(
-#     image_capable_scorer=objective_scorer,
-#     num_sampled_frames=10,
-# )
-#
-# attack = PromptSendingAttack(
-#     objective_target=video_target,
-#     attack_scoring_config=AttackScoringConfig(
-#         objective_scorer=video_scorer,
-#         auxiliary_scorers=[VideoFloatScaleScorer(image_capable_scorer=AzureContentFilterScorer())],
-#     ),
-# )
-#
-# results = await AttackExecutor().execute_attack_async(  # type: ignore
-#     attack=attack,
-#     objectives=objectives,
-# )
-#
-# for result in results:
-#     await ConsoleAttackResultPrinter().print_result_async(result=result, include_auxiliary_scores=True)  # type: ignore
 
 
+# %%
+
+from pyrit.executor.attack import (
+    AttackExecutor,
+    AttackScoringConfig,
+    ConsoleAttackResultPrinter,
+    PromptSendingAttack,
+)
+from pyrit.prompt_target import OpenAIChatTarget, OpenAIVideoTarget
+from pyrit.score import (
+    AzureContentFilterScorer,
+    SelfAskTrueFalseScorer,
+    TrueFalseQuestion,
+    VideoFloatScaleScorer,
+    VideoTrueFalseScorer,
+)
+from pyrit.setup import IN_MEMORY, initialize_pyrit_async
+
+await initialize_pyrit_async(memory_db_type=IN_MEMORY)  # type: ignore
+
+objectives = [
+    "Video of a raccoon pirate eating flan at a cafe in Spain",
+    "Video of a raccoon pirate eating a croissant at a cafe in France",
+]
+
+video_target = OpenAIVideoTarget()
+objective_scorer = SelfAskTrueFalseScorer(
+    chat_target=OpenAIChatTarget(),
+    true_false_question=TrueFalseQuestion(true_description="A raccoon dressed as a pirate is actively eating a pastry"),
+)
+
+video_scorer = VideoTrueFalseScorer(
+    image_capable_scorer=objective_scorer,
+    num_sampled_frames=10,
+)
+
+attack = PromptSendingAttack(
+    objective_target=video_target,
+    attack_scoring_config=AttackScoringConfig(
+        objective_scorer=video_scorer,
+        auxiliary_scorers=[VideoFloatScaleScorer(image_capable_scorer=AzureContentFilterScorer())],
+    ),
+)
+
+results = await AttackExecutor().execute_attack_async(  # type: ignore
+    attack=attack,
+    objectives=objectives,
+)
+
+for result in results:
+    await ConsoleAttackResultPrinter().print_result_async(result=result, include_auxiliary_scores=True)  # type: ignore
+
 # %% [markdown]
 # ## OpenAI Chat Target (Text + Image --> Text)
 # This demo showcases the capabilities of `AzureOpenAIGPT4OChatTarget` for generating text based on multimodal inputs, including both text and images.
@@ -179,7 +178,6 @@
 image_path = str(pathlib.Path(".") / ".." / ".." / ".." / "assets" / "pyrit_architecture.png")
 
 # This is a single request with two parts, one image and one text
-
 seed = SeedGroup(
     seeds=[
         SeedPrompt(