bump: version 1.7.13 → 1.7.14

litellm/main.py

@@ -2212,6 +2212,7 @@ def stream_chunk_builder(chunks: list, messages: Optional[list]=None):
     try:
         response["usage"]["prompt_tokens"] = token_counter(model=model, messages=messages)
     except: # don't allow this failing to block a complete streaming response from being returned
+        print_verbose(f"token_counter failed, assuming prompt tokens is 0")
         response["usage"]["prompt_tokens"] = 0
     response["usage"]["completion_tokens"] = token_counter(model=model, text=completion_output)
     response["usage"]["total_tokens"] = response["usage"]["prompt_tokens"] + response["usage"]["completion_tokens"]

litellm/utils.py

@@ -1613,6 +1613,7 @@ def token_counter(model="", text=None,  messages: Optional[List] = None):
     # use tiktoken, anthropic, cohere or llama2's tokenizer depending on the model
     if text == None:
         if messages is not None:
+            print_verbose(f"token_counter messages received: {messages}")
             text = "".join([message["content"] for message in messages])
         else:
             raise ValueError("text and messages cannot both be None")

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "litellm"
-version = "1.7.13"
+version = "1.7.14"
 description = "Library to easily interface with LLM API providers"
 authors = ["BerriAI"]
 license = "MIT License"
@@ -27,7 +27,7 @@ requires = ["poetry-core"]
 build-backend = "poetry.core.masonry.api"
 
 [tool.commitizen]
-version = "1.7.13"
+version = "1.7.14"
 version_files = [
     "pyproject.toml:^version"
 ]