🧮The LLM Evaluation Framework
active 2025-03-29 → 2026-08-10 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 500 days from 2025-03-29 to 2026-08-10. Pushes: 1,167 total, peak 27 in a day. Pull requests: 642 total, peak 17 in a day. Issues: 225 total, peak 8 in a day. Comments: 921 total, peak 16 in a day. Stars: 4,219 total, peak 161 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| penguine-ip | 775 | 288 | 216 | 197 |
| ji21 | 384 | 384 | 0 | 0 |
| kritinv | 349 | 260 | 61 | 22 |
| vercel[bot] | 328 | 0 | 0 | 328 |
| spike-spiegel-21 | 273 | 169 | 79 | 17 |
| trevor-cai | 124 | 39 | 13 | 40 |
| A-Vamshi | 105 | 3 | 76 | 13 |
| greptile-apps[bot] | 77 | 0 | 0 | 62 |
| BloggerBust | 59 | 0 | 50 | 9 |
| delve-auditor[bot] | 22 | 0 | 0 | 22 |
| Sai-Suraj-27 | 9 | 0 | 9 | 0 |
| ps2program | 9 | 0 | 1 | 6 |
| Stu-ops | 8 | 0 | 2 | 6 |
| karankulshrestha | 7 | 0 | 5 | 2 |
| bofenghuang | 7 | 0 | 2 | 5 |
| JiwaniZakir | 6 | 0 | 1 | 5 |
| tanayvaswani | 5 | 0 | 5 | 0 |
| SYED-M-HUSSAIN | 5 | 0 | 2 | 3 |
| AbhishekRP2002 | 5 | 0 | 1 | 4 |
| uahmad235 | 5 | 0 | 0 | 3 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#2892vercel[bot]2026-07-15 08:08feat: add TokenBudgetMetric — deterministic token/cost budget gate for agent traces
- Issue comment#2300Saswatsusmoy2026-07-14 23:31Feat request: penalize_ambiguous_claims for answer relevancy
- Issue comment#2858vercel[bot]2026-07-07 09:54Voice evals V1
- Issue comment#2643Fame5102026-07-04 23:14Feature: AgentLoopDetectionMetric — detect infinite loops and cyclical tool-call patterns in agent traces
- Pull request#2814luochen2112026-06-28 13:01
- Issue comment#2812xr8432026-06-28 01:47feat(metrics): add AdversarialRobustnessMetric (RoMA-based) (#2150)
- Pull request#2812xr8432026-06-27 04:08
- Pull request#2794k-dickinson2026-06-23 00:55
- Pull request#2792bongho2026-06-20 23:50
- Issue comment#2760Goutham-Annem2026-06-16 23:19docs: add migration note for deprecated ConversationRelevancyMetric
- Pull request#2755atenni2026-06-15 01:37
- Pull request#2753kimnamu2026-06-14 16:04
- Issue#2748hienhtphan2026-06-13 05:49Add retail support evaluation example using synthetic customer support cases
- Issue comment#2721StressTestor2026-06-02 08:15fix(dataset): serialize RetrievedContextData in save_as instead of crashing
- Issue comment#2717vercel[bot]2026-06-01 08:08feat(metric data): add tokens usage
- Issue#1420penguine-ip2026-05-31 04:27Custom LLM in the docs provide no schema input for generate method, and returns string, which causes summarization.py raise AttributeError: 'str' object has no attribute 'truths'
- Pull request#2677zhang-liz2026-05-31 04:27
- Pull request#2713nuthalapativarun2026-05-29 21:32
- Pull request#2711nuthalapativarun2026-05-29 21:29
- Issue comment#2689vercel[bot]2026-05-23 17:58docs: fix typos in tutorials and docs
- Issue comment#2626akshay-kumar-bm2026-05-23 11:41Add OpenRouter Integration for Metrics Evaluation
- Issue comment#947jbbqqf2026-05-23 09:04Example Colab notebook is failing. No module named 'deepeval.metrics.conversation_relevancy
- Issue comment#1262jbbqqf2026-05-23 09:04metric.evaluate() raise TypeError: unsupported operand type(s) for +=: 'NoneType' and 'float'
- Issue comment#1390jbbqqf2026-05-23 09:04missing 1 required positional argument: 'schema' for DAGMetric when using custom LLM
- Issue comment#2676vercel[bot]2026-05-16 23:13feat: add ProcessIntegrityMetric for agent trajectory logical consistency
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 4,219 stars here means stars gained during the window, not the repo's star count.