acreate thin-client

litellm/__init__.py

@@ -4,6 +4,8 @@ input_callback: List[str] = []
 success_callback: List[str] = []
 failure_callback: List[str] = []
 set_verbose = False
+debugger = False # Learn more - https://docs.litellm.ai/docs/debugging/hosted_debugging
+email = None # for debugging dashboard
 telemetry = True
 max_tokens = 256  # OpenAI Defaults
 retry = True
@@ -20,7 +22,6 @@ vertex_location: Optional[str] = None
 togetherai_api_key: Optional[str] = None
 caching = False
 caching_with_models = False # if you want the caching key to be model + prompt
-debugger = False
 model_cost = {
     "gpt-3.5-turbo": {
         "max_tokens": 4000,
@@ -262,7 +263,8 @@ from .utils import (
     cost_per_token,
     completion_cost,
     get_litellm_params,
-    Logging
+    Logging,
+    acreate
 )
 from .main import *  # type: ignore
 from .integrations import *

litellm/integrations/litedebugger.py

@@ -3,14 +3,14 @@ import requests, traceback, json, os
 class LiteDebugger:
     user_email = None
     dashboard_url = None
-    def __init__(self):
+    def __init__(self, email=None):
         self.api_url = "https://api.litellm.ai/debugger"
-        self.validate_environment()
+        self.validate_environment(email)
         pass
 
-    def validate_environment(self):
+    def validate_environment(self, email):
         try:
-            self.user_email = os.getenv("LITELLM_EMAIL")
+            self.user_email = os.getenv("LITELLM_EMAIL") or email
             self.dashboard_url = 'https://admin.litellm.ai/' + self.user_email
             print(f"Here's your free Dashboard 👉 {self.dashboard_url}")
             if self.user_email == None:

litellm/tests/test_async_fn.py

@@ -9,7 +9,7 @@ import asyncio
 sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system path
-from litellm import acompletion
+from litellm import acompletion, acreate
 
 
 async def test_get_response():
@@ -24,3 +24,16 @@ async def test_get_response():
 
 response = asyncio.run(test_get_response())
 print(response)
+
+# async def test_get_response():
+#     user_message = "Hello, how are you?"
+#     messages = [{"content": user_message, "role": "user"}]
+#     try:
+#         response = await acreate(model="gpt-3.5-turbo", messages=messages)
+#     except Exception as e:
+#         pytest.fail(f"error occurred: {e}")
+#     return response
+
+
+# response = asyncio.run(test_get_response())
+# print(response)

litellm/utils.py

@@ -718,7 +718,7 @@ def set_callbacks(callback_list):
                 supabaseClient = Supabase()
             elif callback == "lite_debugger":
                 print(f"instantiating lite_debugger")
-                liteDebuggerClient = LiteDebugger()
+                liteDebuggerClient = LiteDebugger(email=litellm.email)
     except Exception as e:
         raise e
 
@@ -1008,6 +1008,9 @@ def handle_success(args, kwargs, result, start_time, end_time):
         pass
 
 
+def acreate(*args, **kwargs): ## Thin client to handle the acreate langchain call
+    return litellm.acompletion(*args, **kwargs)
+
 def prompt_token_calculator(model, messages):
     # use tiktoken or anthropic's tokenizer depending on the model
     text = " ".join(message["content"] for message in messages)
@@ -1022,7 +1025,6 @@ def prompt_token_calculator(model, messages):
         num_tokens = len(encoding.encode(text))
     return num_tokens
 
-
 def valid_model(model):
     try:
         # for a given model name, check if the user has the right permissions to access the model 

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "litellm"
-version = "0.1.446"
+version = "0.1.447"
 description = "Library to easily interface with LLM API providers"
 authors = ["BerriAI"]
 license = "MIT License"