add factory test

requirements.txt

@@ -3,16 +3,14 @@ langchain-chroma==0.1.4
 langchain_openai==0.2.5
 langgraph==0.2.44
 langchain_community==0.3.5
+langchain_ollama==0.2.0
 transformers==4.45.0
 spacy[ja]==3.8.2
-# --index-url https://download.pytorch.org/whl/cu121
-# torch==2.5.1+cu121
 torch==2.4.0
 torchaudio==2.4.0
-# TTS==0.22.0
+librosa==0.10.1
 git+https://github.com/neoguojing/TTS.git
 ctranslate2==4.4.0
-# nvidia-cudnn-cu12==8.*
 faster-whisper==1.0.3
 diffusers==0.21.2
 accelerate==1.1.1

tests/llms/test_model_factory.py

@@ -1,20 +1,53 @@
 import unittest
 from agi.llms.model_factory import ModelFactory
+from agi.llms.base import build_multi_modal_message,ImageType,AudioType
+from langchain_core.messages import AIMessage, HumanMessage
+
 
 class TestModelFactory(unittest.TestCase):
 
     def test_get_model(self):
         ollama_model = ModelFactory.get_model("ollama")
         resp = ollama_model.invoke("介绍下美国")
-        print(type(resp))
+        self.assertIsNotNone(resp.content)
+        self.assertEqual(len(ModelFactory._instances),1)
+        ModelFactory.destroy("ollama")
+        self.assertEqual(len(ModelFactory._instances),0)
+
+        instance = ModelFactory.get_model("text2image")
+        input = HumanMessage(content="a chinese leader") 
+        resp = instance.invoke(input)
+        self.assertIsNotNone(resp.content)
+        self.assertEqual(len(ModelFactory._instances),1)
+        ModelFactory.destroy("text2image")
+        self.assertEqual(len(ModelFactory._instances),0)
+
+        instance = ModelFactory.get_model("image2image")
+        input = build_multi_modal_message("as a cat woman","tests/cat.jpg",ImageType.FILE_PATH) 
+        resp = instance.invoke(input)
+        self.assertIsNotNone(resp.content)
+        self.assertEqual(len(ModelFactory._instances),1)
+        ModelFactory.destroy("image2image")
+        self.assertEqual(len(ModelFactory._instances),0)
+
+        instance = ModelFactory.get_model("speech2text")
+        input = self.input = build_multi_modal_message("","tests/1730604079.wav",AudioType.FILE_PATH)
+        resp = instance.invoke(input)
+        self.assertIsNotNone(resp.content)
+        print(resp.content)
+        self.assertEqual(len(ModelFactory._instances),1)
+        ModelFactory.destroy("speech2text")
+        self.assertEqual(len(ModelFactory._instances),0)
+
+        instance = ModelFactory.get_model("text2speech")
+        input = HumanMessage(content="岁的思考的加快速度为空军党委科技")
+        resp = instance.invoke(input)
+        self.assertIsNotNone(resp.content)
+        print(resp.content)
+        self.assertEqual(len(ModelFactory._instances),1)
+        ModelFactory.destroy("text2speech")
+        self.assertEqual(len(ModelFactory._instances),0)
 
-    # def test_release_model(self):
-    #     self.assertIsNotNone(output)
-    #     self.assertIsNotNone(output.image)
-    #     self.assertIsNotNone(output.image.pil_image)
-    #     self.assertIsNotNone(output.content)
-    #     print(output.content)
-
 
 if __name__ == "__main__":
     unittest.main()