carlrobertoh / llm-client Goto Github PK

OpenAI has updated gpt 4 turbo preview and now has an alias for their latest model, "gpt-4-turbo-preview" which now points to gpt-4-0125-preview; see docs

Ideally, in CodeGPT preferences you could specify a model name and label instead of having them only hardcoded in OpenAIChatCompletionModel.java

https://github.com/carlrobertoh/LLM-Client/blob/9e6fe09a719a42744a9c8e574180615aa624d4df/src/main/java/ee/carlrobert/llm/client/openai/completion/OpenAIChatCompletionEventSourceListener.java#L19

SSE can come with meta information, which has empty choices array. E.g.:
data: {"id":"","object":"","created":0,"model":"","prompt_filter_results":[{"prompt_index":0,"content_filter_results":{"hate":{"filtered":false,"severity":"safe"},"self_harm":{"filtered":false,"severity":"safe"},"sexual":{"filtered":false,"severity":"safe"},"violence":{"filtered":false,"severity":"safe"}}}],"choices":[]}

Hello,

I am wondering if the following code would not be the cause of my issue when trying to use CodeGPT Remote server (localhost) with LM Studio, which give me the following error (last line) :

[2024-03-29 10:12:29.205] [INFO] [LM STUDIO SERVER] Verbose server logs are ENABLED
[2024-03-29 10:12:29.208] [INFO] [LM STUDIO SERVER] Success! HTTP server listening on port 1234
[2024-03-29 10:12:29.208] [INFO] [LM STUDIO SERVER] Supported endpoints:
[2024-03-29 10:12:29.208] [INFO] [LM STUDIO SERVER] -> GET http://localhost:1234/v1/models
[2024-03-29 10:12:29.209] [INFO] [LM STUDIO SERVER] -> POST http://localhost:1234/v1/chat/completions
[2024-03-29 10:12:29.209] [INFO] [LM STUDIO SERVER] -> POST http://localhost:1234/v1/completions

[2024-03-29 10:12:46.517] [ERROR] Unexpected endpoint or method. (POST /v1/completion). Returning 200 anyway

The other API does use completions (plural), so I am wondering with which server completion (singular) is used ?

Thank you