Merge branch 'main' into docstrings-linting

.github/workflows/tests.yml

@@ -28,6 +28,7 @@ env:
   AZURE_OPENAI_API_KEY: ${{ secrets.AZURE_OPENAI_API_KEY }}
   AZURE_OPENAI_ENDPOINT: ${{ secrets.AZURE_OPENAI_ENDPOINT }}
   GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+  HF_API_TOKEN: ${{ secrets.HUGGINGFACE_API_KEY }}
   PYTHON_VERSION: "3.8"
   HATCH_VERSION: "1.9.3"
 

test/components/generators/chat/test_hugging_face_local.py

@@ -1,12 +1,12 @@
-from unittest.mock import patch, Mock
-from haystack.utils.auth import Secret
+from unittest.mock import Mock, patch
 
 import pytest
 from transformers import PreTrainedTokenizer
 
 from haystack.components.generators.chat import HuggingFaceLocalChatGenerator
 from haystack.dataclasses import ChatMessage, ChatRole
 from haystack.utils import ComponentDevice
+from haystack.utils.auth import Secret
 
 
 # used to test serialization of streaming_callback
@@ -160,7 +160,8 @@ def test_from_dict(self, model_info_mock):
         assert generator_2.streaming_callback is streaming_callback_handler
 
     @patch("haystack.components.generators.chat.hugging_face_local.pipeline")
-    def test_warm_up(self, pipeline_mock):
+    def test_warm_up(self, pipeline_mock, monkeypatch):
+        monkeypatch.delenv("HF_API_TOKEN", raising=False)
         generator = HuggingFaceLocalChatGenerator(
             model="mistralai/Mistral-7B-Instruct-v0.2",
             task="text2text-generation",

test/components/rankers/test_sentence_transformers_diversity.py

@@ -198,10 +198,12 @@ def test_run_without_warm_up(self, similarity):
             ranker.run(query="test query", documents=documents)
 
     @pytest.mark.parametrize("similarity", ["dot_product", "cosine"])
-    def test_warm_up(self, similarity):
+    def test_warm_up(self, similarity, monkeypatch):
         """
         Test that ranker loads the SentenceTransformer model correctly during warm up.
         """
+        monkeypatch.delenv("HF_API_TOKEN", raising=False)
+
         mock_model_class = MagicMock()
         mock_model_instance = MagicMock()
         mock_model_class.return_value = mock_model_instance

test/components/rankers/test_transformers_similarity.py

@@ -1,13 +1,13 @@
+import logging
 from unittest.mock import MagicMock, patch
-from haystack.utils.auth import Secret
 
 import pytest
-import logging
 import torch
 from transformers.modeling_outputs import SequenceClassifierOutput
 
 from haystack import ComponentError, Document
 from haystack.components.rankers.transformers_similarity import TransformersSimilarityRanker
+from haystack.utils.auth import Secret
 from haystack.utils.device import ComponentDevice, DeviceMap
 
 
@@ -272,7 +272,8 @@ def test_device_map_and_device_raises(self, caplog):
 
     @patch("haystack.components.rankers.transformers_similarity.AutoTokenizer.from_pretrained")
     @patch("haystack.components.rankers.transformers_similarity.AutoModelForSequenceClassification.from_pretrained")
-    def test_device_map_dict(self, mocked_automodel, mocked_autotokenizer):
+    def test_device_map_dict(self, mocked_automodel, _mocked_autotokenizer, monkeypatch):
+        monkeypatch.delenv("HF_API_TOKEN", raising=False)
         ranker = TransformersSimilarityRanker("model", model_kwargs={"device_map": {"layer_1": 1, "classifier": "cpu"}})
 
         class MockedModel:

test/components/readers/test_extractive.py

@@ -410,7 +410,8 @@ def __init__(self):
 
 @patch("haystack.components.readers.extractive.AutoTokenizer.from_pretrained")
 @patch("haystack.components.readers.extractive.AutoModelForQuestionAnswering.from_pretrained")
-def test_device_map_auto(mocked_automodel, mocked_autotokenizer):
+def test_device_map_auto(mocked_automodel, _mocked_autotokenizer, monkeypatch):
+    monkeypatch.delenv("HF_API_TOKEN", raising=False)
     reader = ExtractiveReader("deepset/roberta-base-squad2", model_kwargs={"device_map": "auto"})
     auto_device = ComponentDevice.resolve_device(None)
 
@@ -427,7 +428,8 @@ def __init__(self):
 
 @patch("haystack.components.readers.extractive.AutoTokenizer.from_pretrained")
 @patch("haystack.components.readers.extractive.AutoModelForQuestionAnswering.from_pretrained")
-def test_device_map_str(mocked_automodel, mocked_autotokenizer):
+def test_device_map_str(mocked_automodel, _mocked_autotokenizer, monkeypatch):
+    monkeypatch.delenv("HF_API_TOKEN", raising=False)
     reader = ExtractiveReader("deepset/roberta-base-squad2", model_kwargs={"device_map": "cpu:0"})
 
     class MockedModel:
@@ -443,7 +445,8 @@ def __init__(self):
 
 @patch("haystack.components.readers.extractive.AutoTokenizer.from_pretrained")
 @patch("haystack.components.readers.extractive.AutoModelForQuestionAnswering.from_pretrained")
-def test_device_map_dict(mocked_automodel, mocked_autotokenizer):
+def test_device_map_dict(mocked_automodel, _mocked_autotokenizer, monkeypatch):
+    monkeypatch.delenv("HF_API_TOKEN", raising=False)
     reader = ExtractiveReader(
         "deepset/roberta-base-squad2", model_kwargs={"device_map": {"layer_1": 1, "classifier": "cpu"}}
     )

test/components/routers/test_zero_shot_text_router.py

@@ -23,7 +23,8 @@ def test_to_dict(self):
             },
         }
 
-    def test_from_dict(self):
+    def test_from_dict(self, monkeypatch):
+        monkeypatch.delenv("HF_API_TOKEN", raising=False)
         data = {
             "type": "haystack.components.routers.zero_shot_text_router.TransformersZeroShotTextRouter",
             "init_parameters": {