Flight Log / repository
stanford-crfm/levanter — agent flight log
71 pull requests opened by AI coding agents in stanford-crfm/levanter (★ 722, Python): OpenAI Codex (OpenAI) 71 (79% merged).
Flights by agent
| Agent | Flights | Merge rate | Median to merge | Burn rate |
|---|---|---|---|---|
| OpenAI Codex (OpenAI) | 71 | 79% | 24 min | 2% |
By month
| Month | Flights | Merged | Burns |
|---|---|---|---|
| 2025-06 | 11 | 8 | 0 |
| 2025-07 | 60 | 38 | 1 |
Last 50 flights
| PR | Title | Agent | Opened | Lines | |
|---|---|---|---|---|---|
| #1115 | Fix lockfile issue in ray TPU | OpenAI Codex | Stale | Jul 30, 2025 | +53 −19 |
| #1114 | Use DecodeState for generation tracking | OpenAI Codex | Stale | Jul 29, 2025 | +111 −224 |
| #1112 | Implement DecodeState operations | OpenAI Codex | Stale | Jul 28, 2025 | +206 −31 |
| #1111 | Fix GPT-2 integration workflow install | OpenAI Codex | Merged clean | Jul 27, 2025 | +3 −1 |
| #1110 | Fix GPT-2 integration workflow install | OpenAI Codex | Merged clean | Jul 26, 2025 | +1 −1 |
| #1109 | Fix TPU test skipping logic | OpenAI Codex | Merged clean | Jul 26, 2025 | +6 −8 |
| #1108 | Fix docs by re-exporting SkipStepConfig | OpenAI Codex | Merged clean | Jul 26, 2025 | +3 −0 |
| #1106 | Add mixture dataset cache subsampling utility | OpenAI Codex | Stale | Jul 26, 2025 | +149 −0 |
| #1105 | Fix equinox static_field warnings in tests | OpenAI Codex | Abandoned | Jul 26, 2025 | +30 −17 |
| #1102 | Optimize careless LM evaluation | OpenAI Codex | Stale | Jul 26, 2025 | +59 −23 |
| #1099 | Update AGENTS: clarify jit safety rules | OpenAI Codex | Merged clean | Jul 25, 2025 | +10 −0 |
| #1098 | Refactor JitScheduler buffer shape | OpenAI Codex | Merged clean | Jul 25, 2025 | +185 −186 |
| #1096 | Add TPU transfer benchmark script | OpenAI Codex | Stale | Jul 23, 2025 | +4,700 −4,566 |
| #1095 | Add torch-free radix cache | OpenAI Codex | Stale | Jul 23, 2025 | +502 −75 |
| #1091 | Implement Multi-head Latent Attention (MLA) option | OpenAI Codex | Stale | Jul 23, 2025 | +753 −378 |
| #1090 | Update agent planning guideline | OpenAI Codex | Burn | Jul 23, 2025 | +3,159 −2,061 |
| #1088 | Fix trainer missing info error | OpenAI Codex | Merged clean | Jul 22, 2025 | +6 −0 |
| #1087 | Update CI to sync without lock and add lockfile check | OpenAI Codex | Merged clean | Jul 21, 2025 | +26 −3 |
| #1082 | Freeze dataclasses in text.py | OpenAI Codex | Merged clean | Jul 21, 2025 | +10 −10 |
| #1081 | Freeze LayerNormConfigBase classes | OpenAI Codex | Merged clean | Jul 21, 2025 | +4 −4 |
| #1080 | Freeze optimizer dataclasses | OpenAI Codex | Merged clean | Jul 21, 2025 | +26 −26 |
| #1077 | Add dataset token count callback | OpenAI Codex | Stale | Jul 19, 2025 | +89 −24 |
| #1075 | Fix JitScheduler sequence packing | OpenAI Codex | Stale | Jul 19, 2025 | +92 −7 |
| #1074 | Add batch sample lm script | OpenAI Codex | Abandoned | Jul 19, 2025 | +218 −0 |
| #1073 | Fix setup script commit checkout | OpenAI Codex | Merged clean | Jul 19, 2025 | +5 −1 |
| #1072 | Remove redundant test requirements | OpenAI Codex | Abandoned | Jul 19, 2025 | +3,349 −3,825 |
| #1071 | Simplify ray_tpu resource scheduling | OpenAI Codex | Stale | Jul 18, 2025 | +3,689 −3,976 |
| #1070 | Make rotary config fields static | OpenAI Codex | Merged clean | Jul 18, 2025 | +3 −3 |
| #1069 | Add GPU determinism note | OpenAI Codex | Merged clean | Jul 17, 2025 | +3 −0 |
| #1067 | Complete jit scheduler | OpenAI Codex | Merged clean | Jul 17, 2025 | +157 −57 |
| #1066 | Optimize page table slicing | OpenAI Codex | Merged clean | Jul 16, 2025 | +27 −3 |
| #1065 | Optimize PageTable allocation loop | OpenAI Codex | Merged clean | Jul 15, 2025 | +12 −1 |
| #1064 | compute pos_ids during allocation | OpenAI Codex | Merged clean | Jul 15, 2025 | +24 −36 |
| #1062 | Update JittedScheduler to avoid deprecated outfeed | OpenAI Codex | Merged clean | Jul 15, 2025 | +387 −0 |
| #1060 | Add pytree packing utility | OpenAI Codex | Merged clean | Jul 14, 2025 | +132 −51 |
| #1059 | Fix TPU ragged_paged_attention head flattening | OpenAI Codex | Merged clean | Jul 13, 2025 | +15 −4 |
| #1058 | Update sharding helper to avoid deprecated JAX API | OpenAI Codex | Merged clean | Jul 13, 2025 | +14 −11 |
| #1057 | Fix ragged_paged_attention default soft_cap | OpenAI Codex | Merged clean | Jul 13, 2025 | +1 −1 |
| #1056 | Improve sample_lm generation | OpenAI Codex | Abandoned | Jul 13, 2025 | +44 −15 |
| #1055 | Avoid network downloads in tests | OpenAI Codex | Merged clean | Jul 13, 2025 | +35 −13 |
| #1053 | Update CI to use uv best practices | OpenAI Codex | Merged clean | Jul 11, 2025 | +61 −41 |
| #1052 | Fix sampling step after prefill | OpenAI Codex | Merged clean | Jul 11, 2025 | +23 −15 |
| #1051 | Remove KvPageState wrapper | OpenAI Codex | Merged clean | Jul 11, 2025 | +84 −155 |
| #1049 | Add text sampling script | OpenAI Codex | Merged clean | Jul 10, 2025 | +129 −0 |
| #1048 | Move PageTable to dedicated module | OpenAI Codex | Merged clean | Jul 10, 2025 | +202 −211 |
| #1047 | Add helper for ragged paged attention | OpenAI Codex | Merged clean | Jul 10, 2025 | +65 −36 |
| #1045 | Add additional Llama paged decode tests | OpenAI Codex | Merged clean | Jul 10, 2025 | +207 −0 |
| #1043 | Fix Slurm get_process_count | OpenAI Codex | Merged clean | Jul 8, 2025 | +28 −1 |
| #1040 | Add paged decode helper to Llama transformer | OpenAI Codex | Abandoned | Jul 7, 2025 | +112 −26 |
| #1039 | Switch default model to Llama | OpenAI Codex | Abandoned | Jul 7, 2025 | +149 −146 |
No rider has claimed a flight in this repository yet. Flew one of these? Open your flight page and claim it.
Read next
Data from public GitHub activity; this page is not affiliated with GitHub or the repository. Repository owner? Request exclusion — handled within 48 hours.