fix: bumped package versions and improved error handling in notebooks (…

…#184)

dial-cookbook/examples/how_to_call_image_to_text_applications.ipynb

@@ -8,8 +8,8 @@
     "\n",
     "From this notebook, you will learn how to call image-to-text DIAL applications via [DIAL API chat/completions](https://epam-rail.com/dial_api#/paths/~1openai~1deployments~1%7BDeployment%20Name%7D~1chat~1completions/post) call.\n",
     "\n",
-    "\n",    
-    "https://github.com/epam/ai-dial/blob/main/dial-cookbook/examples/how_to_call_image_to_text_applications.ipynb",
+    "\n",
+    "https://github.com/epam/ai-dial/blob/main/dial-cookbook/examples/how_to_call_image_to_text_applications.ipynb\n",
     "\n",
     "\n",
     "**DIAL application** is a general term, which encompasses model adapters and application with any custom logic.\n",
@@ -36,9 +36,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install requests==2.31.0\n",
-    "!pip install openai==1.9.0\n",
-    "!pip install langchain-openai==0.0.3"
+    "!pip install requests==2.32.3\n",
+    "!pip install openai==1.43.0\n",
+    "!pip install langchain-openai==0.1.23"
    ]
   },
   {
@@ -579,7 +579,7 @@
    "source": [
     "## Using LangChain\n",
     "\n",
-    "The [LangChain](https://pypi.org/project/langchain-openai/) library **is not suitable** as a client of image-to-text applications, since `langchain-openai<=0.0.2` ignores the additional fields in the request."
+    "The [LangChain](https://pypi.org/project/langchain-openai/) library **is not suitable** as a client for image-to-text applications, since `langchain-openai` ignores additional fields attached to chat messages in the request."
    ]
   },
   {
@@ -617,8 +617,8 @@
     "  llm.generate(messages=[[HumanMessage(content=\"\", additional_kwargs=extra_fields)]])\n",
     "\n",
     "  raise Exception(\"Generation didn't fail\")\n",
-    "except Exception as e:\n",
-    "  assert str(e) == \"Error code: 422 - {'error': {'message': 'No image attachment was found in the last message', 'type': 'runtime_error'}}\", \"Unexpected error\""
+    "except openai.UnprocessableEntityError as e:\n",
+    "  assert e.body[\"message\"] == \"No image attachment was found in the last message\""
    ]
   },
   {
@@ -634,15 +634,14 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "\n",
     "try:\n",
     "    output = llm.stream(input=[HumanMessage(content=\"\", additional_kwargs=extra_fields)])\n",
     "    for chunk in output:\n",
     "        print(chunk.dict())\n",
     "\n",
     "    raise Exception(\"Generation didn't fail\")\n",
-    "except Exception as e:\n",
-    "    assert str(e) == \"Error code: 422 - {'error': {'message': 'No image attachment was found in the last message', 'type': 'runtime_error'}}\", \"Unexpected error\"\n"
+    "except openai.UnprocessableEntityError as e:\n",
+    "  assert e.body[\"message\"] == \"No image attachment was found in the last message\""
    ]
   }
  ],

dial-cookbook/examples/how_to_call_text_to_image_applications.ipynb

@@ -8,8 +8,8 @@
     "\n",
     "From this notebook, you will learn how to call text-to-image applications via [DIAL API chat/completions](https://epam-rail.com/dial_api#/paths/~1openai~1deployments~1%7BDeployment%20Name%7D~1chat~1completions/post) call.\n",
     "\n",
-    "\n",    
-    "https://github.com/epam/ai-dial/blob/main/dial-cookbook/examples/how_to_call_text_to_image_applications.ipynb",
+    "\n",
+    "https://github.com/epam/ai-dial/blob/main/dial-cookbook/examples/how_to_call_text_to_image_applications.ipynb\n",
     "\n",
     "\n",
     "**DIAL application** is a general term, which encompasses model adapters and applications with any custom logic.\n",
@@ -43,9 +43,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install requests==2.31.0\n",
-    "!pip install openai==1.9.0\n",
-    "!pip install langchain-openai==0.0.3"
+    "!pip install requests==2.32.3\n",
+    "!pip install openai==1.43.0\n",
+    "!pip install langchain-openai==0.1.23"
    ]
   },
   {
@@ -709,7 +709,7 @@
    "source": [
     "## Using LangChain\n",
     "\n",
-    "The [LangChain](https://pypi.org/project/langchain-openai/) library **is not suitable** as a client of text-to-image applications, since `langchain-openai<=0.0.2` ignores the additional fields attached to the response message.\n",
+    "The [LangChain](https://pypi.org/project/langchain-openai/) library **is not suitable** as a client of text-to-image applications, since `langchain-openai` ignores additional fields attached to the response message.\n",
     "\n",
     "Meaning that the response won't contain any image attachments.\n",
     "It's still possible to use the application completion if an application returns image that way:"

dial-cookbook/examples/how_to_call_text_to_text_applications.ipynb

@@ -8,8 +8,8 @@
     "\n",
     "From this notebook, you can learn how to call text-to-text DIAL applications via [DIAL API chat/completions](https://epam-rail.com/dial_api#/paths/~1openai~1deployments~1%7BDeployment%20Name%7D~1chat~1completions/post) call.\n",
     "\n",
-    "\n",    
-    "https://github.com/epam/ai-dial/blob/main/dial-cookbook/examples/how_to_call_text_to_text_applications.ipynb",
+    "\n",
+    "https://github.com/epam/ai-dial/blob/main/dial-cookbook/examples/how_to_call_text_to_text_applications.ipynb\n",
     "\n",
     "\n",
     "For this example, we use a sample text-to-text application called **Echo**, which returns the content of the last user message.\n",
@@ -25,9 +25,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install requests==2.31.0\n",
-    "!pip install openai==1.9.0\n",
-    "!pip install langchain-openai==0.0.3"
+    "!pip install requests==2.32.3\n",
+    "!pip install openai==1.43.0\n",
+    "!pip install langchain-openai==0.1.23"
    ]
   },
   {