Fixes

Author: makseq

.github/workflows/tests.yml

@@ -113,6 +113,7 @@ jobs:
         if: ${{ matrix.skip_pytests != 'true' }}
         env:
           ML_BACKEND: ${{ matrix.backend_dir_name }}
+          TEST_ENV: "true"
         run: |
           docker compose -f label_studio_ml/examples/${{ matrix.backend_dir_name }}/docker-compose.yml exec -T ${{ matrix.backend_dir_name }} pytest -vvv --cov --cov-report=xml:/tmp/coverage.xml
 

label_studio_ml/examples/deepgram/model.py

@@ -1,6 +1,7 @@
 import os
 import sys
 import pathlib
+from types import SimpleNamespace
 from typing import List, Dict, Optional
 from label_studio_ml.model import LabelStudioMLBase
 from label_studio_ml.response import ModelResponse
@@ -20,6 +21,11 @@ class DeepgramModel(LabelStudioMLBase):
 
     def setup(self):
         """Initialize the Deepgram client with API key from environment"""
+        self.test_mode = self._is_test_mode_enabled()
+        if self.test_mode:
+            self._setup_test_clients()
+            return
+
         api_key = os.getenv('DEEPGRAM_API_KEY')
         if not api_key:
             raise ValueError("DEEPGRAM_API_KEY environment variable is not set")
@@ -78,13 +84,46 @@ def predict(self, tasks: List[Dict], context: Optional[Dict] = None, **kwargs) -
             print(f"Uploaded audio to S3: {s3_url}")
 
             # Update task with S3 URL
-            ls.tasks.update(id=task_id, data={"text": text, "audio": s3_url})
+            if self.test_mode:
+                print(f"[TEST MODE] Would update task {task_id} with audio {s3_url}")
+            else:
+                ls.tasks.update(id=task_id, data={"text": text, "audio": s3_url})
         except Exception as e:
             print(f"Error uploading to S3: {e}")
             raise
         finally:
             # Clean up local file
             if os.path.exists(local_audio_path):
                 os.remove(local_audio_path)
-        
+
+    def _is_test_mode_enabled(self) -> bool:
+        """Check environment variables to decide if the model should use local stubs."""
+        truthy = {'1', 'true', 'TRUE', 'True', 'yes', 'on'}
+        explicit_flag = os.getenv('DEEPGRAM_TEST_MODE')
+        test_env_flag = os.getenv('TEST_ENV')
+        return (explicit_flag in truthy) or (test_env_flag in truthy)
+
+    def _setup_test_clients(self):
+        """Configure lightweight stub clients so docker/CI runs do not need real secrets."""
+        print("[TEST MODE] DeepgramModel using stubbed Deepgram/S3 clients.")
+
+        def fake_generate(text: str):
+            # Produce deterministic fake audio bytes for predictable tests.
+            preview = text[:10] if text else ''
+            return [f"fake-audio-{preview}".encode('utf-8')]
+
+        fake_audio = SimpleNamespace(generate=fake_generate)
+        fake_speak = SimpleNamespace(v1=SimpleNamespace(audio=fake_audio))
+        self.deepgram_client = SimpleNamespace(speak=fake_speak)
+
+        class _StubS3Client:
+            """Minimal S3 client replacement for test environments."""
+            def upload_file(self, filename, bucket, key, ExtraArgs=None):
+                print(f"[TEST MODE] Pretend upload of {filename} to s3://{bucket}/{key}")
+
+        self.s3_client = _StubS3Client()
+        # Provide sensible defaults so downstream URL building still works.
+        self.s3_region = os.getenv('AWS_DEFAULT_REGION', 'us-east-1')
+        self.s3_bucket = os.getenv('S3_BUCKET', 'test-bucket')
+        self.s3_folder = os.getenv('S3_FOLDER', 'tts')
 

label_studio_ml/examples/deepgram/test_model.py

@@ -178,3 +178,44 @@ def test_predict_s3_failure_raises_and_cleans_up_temp_file(env_settings, patched
     assert not os.path.exists(local_path)
     patched_clients['ls'].tasks.update.assert_not_called()
 
+
+def test_setup_in_test_mode_uses_stub_clients(monkeypatch):
+    """
+    Scenario: TEST_ENV is enabled to activate internal stubs.
+    Steps   : set TEST_ENV, instantiate the model, and inspect configured clients.
+    Checks  : ensure real Deepgram constructor is not called and stub clients exist with defaults.
+    """
+    monkeypatch.setenv('TEST_ENV', '1')
+    ctor = MagicMock()
+    monkeypatch.setattr(deepgram_model, 'DeepgramClient', ctor)
+
+    model = deepgram_model.DeepgramModel()
+
+    assert model.test_mode is True
+    ctor.assert_not_called()
+    assert callable(model.deepgram_client.speak.v1.audio.generate)
+    assert model.s3_bucket == 'test-bucket'
+    assert model.s3_folder == 'tts'
+
+
+def test_predict_test_mode_skips_label_studio_update(monkeypatch):
+    """
+    Scenario: predict runs in test mode so external Label Studio updates should be skipped.
+    Steps   : enable TEST_ENV, patch ls.tasks.update, run predict with valid context.
+    Checks  : confirm stub S3 upload runs without raising and Label Studio update is not invoked.
+    """
+    monkeypatch.setenv('TEST_ENV', '1')
+    model = deepgram_model.DeepgramModel()
+    mocked_update = MagicMock()
+    monkeypatch.setattr(deepgram_model.ls.tasks, 'update', mocked_update)
+
+    tasks = [{'id': 321}]
+    context = {
+        'user_id': 'tester',
+        'result': [{'value': {'text': ['Hello from test mode']}}],
+    }
+
+    model.predict(tasks=tasks, context=context)
+
+    mocked_update.assert_not_called()
+