Repository navigation
Run the llm eval on Workers AI and publish the results - #22
Merged
Merged
Conversation
The llm row in the README said "not run yet". This runs it on @cf/zai-org/glm-4.7-flash through the free /api/llm path, and the eval script no longer asks for an API key there. I added --offset so a run can be split. It covers the first 10 interests (1,022 neurons) so it stays well inside the shared daily pool. Schema 10/10, structure 10/10, coverage 0.95, on-topic 0.55, and 0/60 citations verified when the model writes its own evidence.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
|
Important
This repository does not receive automatic reviews because it has fewer than 10 stars. ⚙️ Run configuration
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The README's
llmrow has said "not run yet" since the eval was added. This PR runs it.Harness
--mode llmwithLLM_PROVIDER=freenow runs against a Worker's/api/llmwithout needing an API key. It needsLLM_BASE_URL, defaults to@cf/zai-org/glm-4.7-flash, and reportstotalNeuronslikeworkermode does. Other providers behave as before.--offset Nlets you run a slice. Every summary now recordsoffsetandlimit.Run (2026-10-20,
@cf/zai-org/glm-4.7-flash, first 10 of 20 interests)The model writes its own evidence in this mode, and none of its 60 citations checked out: 55 weren't DOI or OpenAlex links, 4 were DOIs that don't exist, and 1 had a mismatched title. So the pipeline holds its shape, but the sources are made up. That's the case for fetching from OpenAlex first.
It cost 1,022.29 neurons (about 102 per run). A full 20 would have come to about 2,100 out of the shared 10k/day pool, so I stopped at 10. The results are two calls with the same settings, merged: a 1-run cost probe, then 9 more. Raw JSON and notes are in
evals/results/2026-10-20-glm-4.7-flash-llm.{json,md}.pnpm typecheck:serverandpnpm test(99) pass locally.