Skip to content

Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most LLMs and agent frameworks including CrewAI, Agno, OpenAI Agents SDK, Langchain, Autogen, AG2, and CamelAI

fixed v1 vs v0 patching

reibs35ca2fd59b0d
diff --git a/agentops/llm_tracker.py b/agentops/llm_tracker.py
index d1ec0e8e..ba7162e9 100644
--- a/agentops/llm_tracker.py
+++ b/agentops/llm_tracker.py
@@ -24,10 +24,8 @@ class LlmTracker:
def __init__(self, client):
self.client = client
self.event_stream = None
- self.override_openai_completion()
- self.override_openai_async_completion()
- def _handle_response_openai(self, response, kwargs, init_timestamp, v1=False):
+ def _handle_response_openai(self, response, kwargs, init_timestamp):
"""Handle responses for OpenAI versions <v1.0.0"""
def handle_stream_chunk(chunk):
try:
@@ -78,38 +76,46 @@ class LlmTracker:
yield chunk
return generator()
- # OpenAI v1.0.0+ response:
- if v1:
- ...
- # self.client.record(Event(
- # event_type=response.object,
- # params=kwargs,
- # result='Success',
- # returns={"content": response.choices[0].message},
- # action_type='llm',
- # model=response.model,
- # prompt=kwargs['messages'],
- # init_timestamp=init_timestamp
- # ))
- # Standard response
- else:
+ # v0.0.0 responses are dicts
+ try:
+ self.client.record(Event(
+ event_type=response['object'],
+ params=kwargs,
+ result='Success',
+ returns={"content":
+ response['choices'][0]['message']['content']},
+ action_type='llm',
+ model=response['model'],
+ prompt=kwargs['messages'],
+ init_timestamp=init_timestamp,
+ prompt_tokens=response.get('usage',
+ {}).get('prompt_tokens'),
+ completion_tokens=response.get('usage',
+ {}).get('completion_tokens')
+ ))
+ except:
+ # v1.0.0+ responses are objects
try:
self.client.record(Event(
- event_type=response['object'],
+ event_type=response.object,
params=kwargs,
result='Success',
- returns={"content":
- response['choices'][0]['message']['content']},
+ returns={
+ # TODO: Will need to make the completion the key for content, splat out the model dump
+ "content": response.choices[0].message.model_dump()},
action_type='llm',
- model=response['model'],
+ model=response.model,
prompt=kwargs['messages'],
- init_timestamp=init_timestamp
+ init_timestamp=init_timestamp,
+ prompt_tokens=response.usage.prompt_tokens,
+ completion_tokens=response.usage.completion_tokens
))
+ # Standard response
except:
print(
f"Unable to parse response for LLM call {kwargs} - skipping upload to AgentOps")
- return response
+ return response
def _handle_async_response_openai(self, response, kwargs, init_timestamp):
def handle_stream_chunk(chunk):
@@ -183,8 +189,7 @@ class LlmTracker:
def _override_method(self, api, method_path, module):
def handle_response(result, kwargs, init_timestamp):
if api == "openai":
- return self._handle_response_openai(
- result, kwargs, init_timestamp)
+ return self._handle_response_openai(result, kwargs, init_timestamp)
return result
def wrap_method(original_method):
@@ -224,10 +229,18 @@ class LlmTracker:
module = import_module(api)
+ # If openai 1.0.0+ is specified, patch the completions methods
+ if api == 'openai':
+ module_version = parse(module.__version__)
+ if module_version >= parse('1.0.0'):
+ self.override_openai_v1_completion()
+ self.override_openai_v1_async_completion()
+
if hasattr(module, '__version__'):
module_version = parse(module.__version__)
for version in sorted(self.SUPPORTED_APIS[api], key=parse, reverse=True):
if module_version >= parse(version):
+ # Patch all methods in every API
for method_path in self.SUPPORTED_APIS[api][version]:
self._override_method(api, method_path, module)
break
@@ -235,7 +248,7 @@ class LlmTracker:
for method_path in self.SUPPORTED_APIS[api]['0.0.0']:
self._override_method(api, method_path, module)
- def override_openai_completion(self):
+ def override_openai_v1_completion(self):
from openai.resources.chat import completions
# Store the original method
@@ -243,20 +256,17 @@ class LlmTracker:
# Define the patched function
def patched_function(*args, **kwargs):
- print('patched function')
- print(args)
- print(kwargs)
+ init_timestamp = get_ISO_time()
# Call the original function with its original arguments
result = original_create(*args, **kwargs)
- self._handle_response_openai(result, kwargs, get_ISO_time())
+ self._handle_response_openai(result, kwargs, init_timestamp)
- # You can add additional logic here if needed
return result
# Override the original method with the patched one
completions.Completions.create = patched_function
- def override_openai_async_completion(self):
+ def override_openai_v1_async_completion(self):
from openai.resources.chat import completions
# Store the original method