Fix LLM AI client not returning a HTTP response with 204 responses (!172025) · Merge requests · GitLab.org / GitLab · GitLab

Snippets Groups Projects

Merged Stan Hu requested to merge sh-fix-llm-ai-client-204-response into master 4 months ago

Nov 12, 2024

Improve HTTP error checking in AiGateway::Completions::Base · 4d0e8965
Stan Hu authored 4 months ago
```
This commit now handles a blank HTTP body and checks that the request
actually succeeded.
```
Verified

4d0e8965

Improve Vertex AI HTTP response handling · 36abedd2

Stan Hu authored 4 months ago

This commit restores the previous behavior of returning `nil` if the
response body is empty and improves the error messages around this.

36abedd2

Nov 11, 2024

Improve error checking in EmbeddingsCompletion · c620e73f

Stan Hu authored 4 months ago

Previously EmbeddingsCompletion did not check the HTTP response status
sufficiently. This commit ensures the responses return 200 and
improves upon the tests.

c620e73f

Explicitly check response status in Anthropic client · 7783651d

Stan Hu authored 4 months ago

Since 500 requests can return `nil`, we need to ensure
that Anthropic requests are checked properly. In addition,
since the body can be blank, preserve the previous behavior
of returning nothing if no body exists.

7783651d

Nov 08, 2024

Fix LLM AI client not returning a HTTP response with 204 responses · 98e578c9

Stan Hu authored 4 months ago

Previously `Llm::AiGateway::Client#request` returned `nil` if the AI
Gateway return a `204 No Content` response. However, this makes it
impossible to discern whether the request was successful or whether
the server returned a 5xx error.

This was happening because `run_retry_with_exponential_backoff`
returned `nil` if the body were blank. To fix this, return the
HTTParty response even the body is blank, but ensure the callers
handle this.

Changelog: fixed

98e578c9