search

aurelio-labs / semantic-router

Superfast AI decision making and intelligent processing of multi-modal data.
https://www.aurelio.ai/semantic-router
MIT License
1.91k stars 201 forks source link

Router stuck in WARNING semantic_router.utils.logger Retrying in 2 seconds #354

Open mvohal opened 2 months ago

mvohal commented 2 months ago

openai | 1.35.13 langchain | 0.2.8 langchain-core | 0.2.19 semantic-router | 0.0.52

`from openai import OpenAI import os os.environ["OPENAI_API_KEY"] = "sk-XXXXX" from semantic_router import Route, RouteLayer from semantic_router.encoders import OpenAIEncoder

repo_route = Route( name="get-repos", utterances=[ "what are the repos in ADSI?", "what are the files in the repo", ], )

project_route = Route( name="get-project-details", utterances=[ "is there a project called ADSI?", "can you give me details around the project ADSI?", ], )

routes=[repo_route, project_route]

rl = RouteLayer(encoder = OpenAIEncoder(name="text-embedding-3-small"),routes=routes) route=rl("list all the files in the repo ADS") print(route.name)`

2024-07-15 15:50:20 WARNING semantic_router.utils.logger Retrying in 2 seconds... 2024-07-15 15:50:27 WARNING semantic_router.utils.logger Retrying in 4 seconds... 2024-07-15 15:50:38 WARNING semantic_router.utils.logger Retrying in 8 seconds... 2024-07-15 15:50:57 WARNING semantic_router.utils.logger Retrying in 16 seconds... 2024-07-15 15:51:31 WARNING semantic_router.utils.logger Retrying in 32 seconds... 2024-07-15 15:52:38 WARNING semantic_router.utils.logger Retrying in 64 seconds... 2024-07-15 15:52:38 INFO semantic_router.utils.logger Returned embeddings: None Traceback (most recent call last): File "/Users/mmvohal/code/ai-thon/pages/📖FOSS.py", line 27, in rl = RouteLayer(encoder = OpenAIEncoder(name="text-embedding-ada-002"),routes=routes) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/mmvohal/code/ai-thon/ai-thon/lib/python3.11/site-packages/semantic_router/layer.py", line 222, in init self._add_routes(routes=self.routes) File "/Users/mmvohal/code/ai-thon/ai-thon/lib/python3.11/site-packages/semantic_router/layer.py", line 465, in _add_routes embedded_utterances = self.encoder(all_utterances) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/mmvohal/code/ai-thon/ai-thon/lib/python3.11/site-packages/semantic_router/encoders/openai.py", line 135, in call raise ValueError(f"No embeddings returned. Error: {error_message}") ValueError: No embeddings returned. Error: Connection error.

jamescalam commented 2 months ago

The warnings you're seeing here are from the retry mechanism for embeddings. The encoder retries if it gets an erroneous (or no) response from OpenAI, likely issues:

  1. Actual network connection issues
  2. API key?
  3. Does your API key have permissions to create embeddings?
mvohal commented 2 months ago

No network connection issues API Key is correct Everything is working as expected on Windows machine but issue is on mac os (M1)

jess-lord commented 1 month ago

i also get this. I was testing cohere/huggingface/openai. OpenAI worked, then later in the day it stopped working giving a similar error to OP. Generated a new API key with no restrictions and same behaviour. 

2024-08-12 14:52:33 WARNING semantic_router.utils.logger Retrying in 2 seconds...
2024-08-12 14:52:38 WARNING semantic_router.utils.logger Retrying in 4 seconds...
Traceback (most recent call last):
...
BadRequestError: Error code: 400 - {'error': {'message': "'$.input' is invalid. Please check the API reference: https://platform.openai.com/docs/api-reference.", 'type': 'invalid_request_error', 'param': None, 'code': None}}