TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and support state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT-LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in performant way.
Python · active 2024-11-30 → 2026-08-11 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 620 days from 2024-11-30 to 2026-08-11. Pushes: 4,303 total, peak 28 in a day. Pull requests: 7,195 total, peak 55 in a day. Issues: 2,208 total, peak 45 in a day. Comments: 75,405 total, peak 496 in a day. Stars: 2,224 total, peak 57 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 |
|---|---|---|---|---|
| tensorrt-cicd | 33.7K | 243 | 23 | 33.4K |
| coderabbitai[bot] | 10.8K | 0 | 0 | 6.7K |
| yuxianq | 1.5K | 83 | 106 | 765 |
| Superjomn | 1.3K | 81 | 143 | 745 |
| lucaslie | 1.3K | 66 | 104 | 601 |
| chzblych | 1.3K | 174 | 161 | 517 |
| github-actions[bot] | 1.2K | 0 | 406 | 428 |
| nv-guomingz | 1.2K | 123 | 203 | 617 |
| QiJune | 1.2K | 121 | 210 | 531 |
| kaiyux | 1.2K | 163 | 196 | 509 |
| brb-nv | 1K | 67 | 102 | 578 |
| EmmaQiaoCh | 869 | 98 | 151 | 516 |
| hyukn | 836 | 53 | 84 | 506 |
| syuoni | 822 | 63 | 92 | 452 |
| chang-l | 792 | 56 | 68 | 454 |
| karljang | 790 | 20 | 38 | 451 |
| xinhe-nv | 770 | 101 | 228 | 349 |
| 2ez4bz | 734 | 55 | 54 | 449 |
| venkywonka | 731 | 49 | 81 | 396 |
| Tabrizian | 722 | 60 | 93 | 404 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#17251tensorrt-cicd2026-08-11 02:58[None][infra] Align VisualGen CBTS rule with CODEOWNERS scope
- Issue comment#11140brnguyen22026-08-10 18:30[TRTLLM-9802][feat] Accelerate L0 torch compile test by reducing num …
- Issue comment#16902tensorrt-cicd2026-08-10 14:41[https://nvbugs/6262973][perf] Move AllReduce autotuner dispatch to C++
- Issue comment#16987tensorrt-cicd2026-08-10 06:23[TRTLLM-14730][feat] Add image edit serving endpoint for visual generation models
- Issue comment#14097trtllm-agent2026-08-09 19:13[https://nvbugs/6162618][fix] Add `extra_acc_spec: tp_attn` reference entry (92.0) in gsm8k.yaml, pass `extra_
- Issue comment#17324chienchunhung2026-08-08 03:00[None][fix] Simplify idle disagg KV transfer progress check
- Issue comment#17374tensorrt-cicd2026-08-08 02:07[None][fix] Use HND mapping for MiniMax-M3 MSA KV cache
- Issue comment#17366tensorrt-cicd2026-08-08 00:48[TRTLLM-14388][refactor] BREAKING: Force 2 model spec dec to fall back to 1 model
- Issue comment#17014tensorrt-cicd2026-08-08 00:46[https://nvbugs/6565412][fix] Size trtllm-gen and thop decode buffers for beam search
- Issue comment#16959tensorrt-cicd2026-08-07 21:51[https://nvbugs/6523820][fix] Waive NCCL-EP dispatch-only CUDA graph replay test
- Issue comment#17076tensorrt-cicd2026-08-07 20:43[https://nvbugs/6533914][docs] Fix trtllm-bench prepare-dataset: --tokenizer is not a valid global option
- Pull request#17430moraxu2026-08-07 20:43
- Issue comment#16162tensorrt-cicd2026-08-07 20:09[None][feat] Add FastWan2.2 TI2V-5B DMD pipeline (3-step text-to-video)
- Issue comment#17366tensorrt-cicd2026-08-07 19:35[TRTLLM-14388][refactor] BREAKING: Force 2 model spec dec to fall back to 1 model
- Issue comment#17375tensorrt-cicd2026-08-07 19:33[https://nvbugs/6428096][fix] Unwaive DeepSeekV3Lite compiled tp2pp2
- Issue comment#17375cascade8122026-08-07 19:27[https://nvbugs/6428096][fix] Unwaive DeepSeekV3Lite compiled tp2pp2
- Issue comment#16849tensorrt-cicd2026-08-07 19:26[None][perf] fp8 block scale quant fusion in SM90 Cutlass MoE
- Issue comment#17364nv-lschneider2026-08-07 19:19[https://nvbugs/6517844][test] Unwaive DeepSeek V3 Lite RTX Pro 6000D test
- Issue comment#16951tensorrt-cicd2026-08-07 19:11[TRTLLM-14692][feat] Overlap LoRA and base model computations
- Issue comment#17412AlessioNetti2026-08-07 19:10[TRTLLM-15192][feat] Specialize CUDA graphs for LoRA
- Issue comment#16951AlessioNetti2026-08-07 19:06[TRTLLM-14692][feat] Overlap LoRA and base model computations
- Issue comment#16914tensorrt-cicd2026-08-07 19:03[None][feat] Support DFlash RoPE, sliding-window configuration, and TRTLLM-gen attention backend
- Issue comment#17191zcxGGmu2026-08-07 18:49fix: guard PRE_MLP NVFP4 fusion for dense layers
- Issue comment#17324tensorrt-cicd2026-08-07 18:38[None][fix] Simplify idle disagg KV transfer progress check
- Issue comment#17191zcxGGmu2026-08-07 18:38fix: guard PRE_MLP NVFP4 fusion for dense layers
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 2,224 stars here means stars gained during the window, not the repo's star count.