Open Moep90 opened 1 year ago
Just to clarify, since I didn't see it in the OP, when using the LocalAI
provider with a base url that includes localhost
, the error shown in the logs is:
INFO 2024-01-12 08:07:41,631 coordinator Actor _inputs1 has no dependencies. Sending BEGIN message
ERROR 2024-01-12 08:07:41,633 __init__ Exception occurred while processing
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/llmstack/common/blocks/llm/__init__.py", line 42, in process
return self._process(self.parse_validate_input(input), self.configuration)
File "/usr/local/lib/python3.10/dist-packages/llmstack/common/blocks/llm/openai.py", line 136, in _process
http_input = HttpAPIProcessorInput(
File "pydantic/main.py", line 341, in pydantic.main.BaseModel.__init__
pydantic.error_wrappers.ValidationError: 1 validation error for HttpAPIProcessorInput
url
URL host invalid, top level domain required (type=value_error.url.host)
Describe the bug Unable to add localAI to docker-compose and testing it since URL requires TLD
To Reproduce
Expected behavior LLMstack should be able to communicate with local local-ai like
http://local-ai:8080
Version v0.0.7
Environment DISTRIB_DESCRIPTION="Linux Mint 21.2 Victoria" Docker version 24.0.5, build ced0996 Docker Compose version v2.20.3
Screenshots
Additional context