create model response object

docs/my-website/docs/completion/output.md

@@ -1,12 +1,20 @@
-# Completion Function - completion()
-Here's the exact json output you can expect from a litellm `completion` call:
+# Output Format - completion()
+Here's the exact json output you can expect from all litellm `completion` calls for all models
 
 ```python 
-{'choices': [{'finish_reason': 'stop',
-   'index': 0,
-   'message': {'role': 'assistant',
-    'content': " I'm doing well, thank you for asking. I am Claude, an AI assistant created by Anthropic."}}],
+{
+  'choices': [
+     {
+        'finish_reason': 'stop',
+        'index': 0,
+        'message': {
+           'role': 'assistant',
+            'content': " I'm doing well, thank you for asking. I am Claude, an AI assistant created by Anthropic."
+        }
+      }
+    ],
  'created': 1691429984.3852863,
  'model': 'claude-instant-1',
- 'usage': {'prompt_tokens': 18, 'completion_tokens': 23, 'total_tokens': 41}}
+ 'usage': {'prompt_tokens': 18, 'completion_tokens': 23, 'total_tokens': 41}
+}
 ```

litellm/main.py

@@ -11,21 +11,10 @@ from .llms.huggingface_restapi import HuggingfaceRestAPILLM
 import tiktoken
 from concurrent.futures import ThreadPoolExecutor
 encoding = tiktoken.get_encoding("cl100k_base")
-from litellm.utils import get_secret, install_and_import, CustomStreamWrapper, read_config_args
+from litellm.utils import get_secret, install_and_import, CustomStreamWrapper, ModelResponse, read_config_args
 from litellm.utils import get_ollama_response_stream, stream_to_string, together_ai_completion_streaming
 ####### ENVIRONMENT VARIABLES ###################
 dotenv.load_dotenv() # Loading env variables using dotenv
-new_response = {
-        "choices": [
-          {
-            "finish_reason": "stop",
-            "index": 0,
-            "message": {
-                "role": "assistant"
-            }
-          }
-        ]
-      }
 # TODO add translations
 ####### COMPLETION ENDPOINTS ################
 #############################################
@@ -54,7 +43,8 @@ def completion(
     top_k=40, request_timeout=0, # unused var for old version of OpenAI API
   ):
   try:
-    global new_response
+    new_response = ModelResponse()
+    print(new_response)
     if azure: # this flag is deprecated, remove once notebooks are also updated.
       custom_llm_provider="azure"
     args = locals()

litellm/tests/test_completion.py

@@ -119,6 +119,14 @@ def test_completion_openai_with_more_optional_params():
         response = completion(model="gpt-3.5-turbo", messages=messages, temperature=0.5, top_p=0.1, n=2, max_tokens=150, presence_penalty=0.5, frequency_penalty=-0.5, logit_bias={123: 5}, user="ishaan_dev@berri.ai")
         # Add any assertions here to check the response
         print(response)
+        response_str = response['choices'][0]['message']['content']
+        response_str_2 = response.choices[0].message.content
+        print(response['choices'][0]['message']['content'])
+        print(response.choices[0].message.content)
+        if type(response_str) != str:
+            pytest.fail(f"Error occurred: {e}")
+        if type(response_str_2) != str:
+            pytest.fail(f"Error occurred: {e}")
     except Exception as e:
         pytest.fail(f"Error occurred: {e}")
 

litellm/utils.py

@@ -30,6 +30,51 @@ user_logger_fn = None
 additional_details = {}
 local_cache = {}
 
+######## Model Response #########################
+# All liteLLM Model responses will be in this format, Follows the OpenAI Format
+# https://docs.litellm.ai/docs/completion/output
+# {
+#   'choices': [
+#      {
+#         'finish_reason': 'stop',
+#         'index': 0,
+#         'message': {
+#            'role': 'assistant',
+#             'content': " I'm doing well, thank you for asking. I am Claude, an AI assistant created by Anthropic."
+#         }
+#       }
+#     ],
+#  'created': 1691429984.3852863,
+#  'model': 'claude-instant-1',
+#  'usage': {'prompt_tokens': 18, 'completion_tokens': 23, 'total_tokens': 41}
+# }
+
+class ModelResponse:
+    def __init__(self):
+        self.choices = [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "message": {
+                "role": "assistant"
+            }
+          }
+        ]
+        self.created = None
+        self.model = None
+        self.usage = {
+            "prompt_tokens": None,
+            "completion_tokens": None,
+            "total_tokens": None
+        }
+
+    def __getitem__(self, key):
+        return getattr(self, key)
+
+    def __setitem__(self, key, value):
+        setattr(self, key, value)
+
+
 def print_verbose(print_statement):
   if litellm.set_verbose:
     print(f"LiteLLM: {print_statement}")