Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[BUG] Wrong GPT version displayed in the trace interface #4717

Open
Howie-Arup opened this issue Sep 23, 2024 · 4 comments
Open

[BUG] Wrong GPT version displayed in the trace interface #4717

Howie-Arup opened this issue Sep 23, 2024 · 4 comments
Assignees
Labels
bug Something isn't working needs information Needs more info from the issuer

Comments

@Howie-Arup
Copy link

Howie-Arup commented Sep 23, 2024

Describe the bug
I am using Phoenix to trace my LLM application locally. My open AI key is GPT 4. But when I open the tracing interface in the browser, it shows that it's GPT 3.5 Turbo, as shown in the screenshot below.

To Reproduce
Steps to reproduce the behavior:

azure_api_key = "..."
azure_endpoint = "..."
azure_api_version = "2024-02-15-preview" # "2023-07-01-preview" "2023-12-01-preview" "2024-02-15-preview"

embed_azure_api_key = "..."
embed_azure_endpoint = "..."
embed_azure_api_version = "2023-12-01-preview" # "2023-07-01-preview" "2023-12-01-preview"


OpenAI_LLM_name = "gpt-4-1106-preview" #'gpt-35-turbo', 'gpt-4' 'gpt-4-1106-preview"
OpenAI_Embedding_model_name = "text-embedding-ada-002"

os.environ["OPENAI_API_VERSION"] = azure_api_version 
os.environ["AZURE_OPENAI_ENDPOINT"] = azure_endpoint
os.environ["AZURE_OPENAI_API_KEY"] = azure_api_key

os.environ["AZURE_OPENAI_VERSION"] = azure_api_version 
os.environ["AZURE_OPENAI_DEPLOYMENT"] = OpenAI_LLM_name
os.environ["AZURE_OPENAI_KEY"] = azure_api_key

# ... Codes of the LLM application 

# Save
directory = 'traces'
trace_id = px.Client().get_trace_dataset().save(directory=directory)

# Load
trace_dataset = px.TraceDataset.load(trace_id, directory=directory)
px.launch_app(trace=trace_dataset)

The after openning the browser, it shows gpt-35-turbo.

Expected behavior
The interface shows gpt 4 correctly.

Screenshots
image

Environment (please complete the following information):

  • OS: Windows 11
  • Notebook Runtime: Jupyter notebook in VS Code
  • Browser: Chrome
  • Version:
arize-phoenix             4.36.0                   pypi_0    pypi
arize-phoenix-evals       0.15.1                   pypi_0    pypi
arize-phoenix-otel        0.5.0                    pypi_0    pypi
@Howie-Arup Howie-Arup added bug Something isn't working triage issues that need triage labels Sep 23, 2024
@github-project-automation github-project-automation bot moved this to 📘 Todo in phoenix Sep 23, 2024
@axiomofjoy
Copy link
Contributor

Thanks for reporting @Howie-Arup. Just to double-check, are you using openinference-instrumentation-openai?

@axiomofjoy axiomofjoy added needs information Needs more info from the issuer and removed triage issues that need triage labels Sep 23, 2024
@axiomofjoy axiomofjoy self-assigned this Sep 23, 2024
@Howie-Arup
Copy link
Author

openinference-instrumentation-openai

Thanks for reply @axiomofjoy. I am not using the openinference-instrumentation-openai library.

@axiomofjoy
Copy link
Contributor

Hey @Howie-Arup, what instrumentation library are you using to create your traces?

@axiomofjoy axiomofjoy assigned axiomofjoy and unassigned axiomofjoy Sep 28, 2024
@Howie-Arup
Copy link
Author

Hi @axiomofjoy I am using LlamaIndex

px.launch_app()
llama_index.core.set_global_handler("arize_phoenix")

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working needs information Needs more info from the issuer
Projects
Status: 📘 Todo
Development

No branches or pull requests

2 participants