Skip to content

Pull requests: NVIDIA/TensorRT-LLM

New pull request New
70 Open 350 Closed
70 Open 350 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Custom samplingconfig addition
#2633 opened Dec 27, 2024 by buddhapuneeth Loading…
Create c-cpp.yml
#2601 opened Dec 20, 2024 by TNGBBK Loading…
Constraint pynvml version
#2570 opened Dec 12, 2024 by MahmoudAshraf97 Loading…
fix NV bench output len was garbage value
#2516 opened Nov 29, 2024 by ekagra-ranjan Loading…
[DO NOT MERGE] test CI
#2513 opened Nov 29, 2024 by niukuo Loading…
1
feat(qwen): add trust_remote_code argument support triaged Issue has been triaged by maintainers
#2493 opened Nov 24, 2024 by ShivamSphn Loading…
bugfix/incorrect lora out dims triaged Issue has been triaged by maintainers
#2484 opened Nov 22, 2024 by akhoroshev Loading…
Fix prompt_table_data empty tensor shape error triaged Issue has been triaged by maintainers
#2470 opened Nov 20, 2024 by BasicCoder Loading…
1
Create INT8 KV Cache on Qserve triaged Issue has been triaged by maintainers
#2446 opened Nov 14, 2024 by dleunji Loading…
2
th::optional -> std::optional triaged Issue has been triaged by maintainers
#2397 opened Oct 31, 2024 by r-barnes Loading…
attention mechanism toggle added functionality issue triaged Issue has been triaged by maintainers waiting for feedback
#2384 opened Oct 28, 2024 by Aaryanverma Loading…
@nv-guomingz
1
fix load_model_on_cpu on qwen/convert_checkpoint.py feature request New feature or request triaged Issue has been triaged by maintainers
#2382 opened Oct 27, 2024 by lkm2835 Loading…
Fix errors when using smoothquant to quantize Qwen2 model Low Precision Issue about lower bit quantization, including int8, int4, fp8 triaged Issue has been triaged by maintainers
#2370 opened Oct 24, 2024 by Missmiaom Loading…
1
README.md: Add 3rd Party Inference Speed Dashboard Documentation Improvements or additions to documentation triaged Issue has been triaged by maintainers
#2244 opened Sep 22, 2024 by matichon-vultureprime Loading…
@laikhtewari
1
Modify small-batched weight only quantization Low Precision Issue about lower bit quantization, including int8, int4, fp8 triaged Issue has been triaged by maintainers
#2213 opened Sep 10, 2024 by dasistwo Loading…
5
[examples/bert/build.py]: Load weights for BertModel and RobertaModel if --model_dir is provided triaged Issue has been triaged by maintainers
#2187 opened Sep 3, 2024 by tkhanipov Loading…
@symphonylyh
4
Create sync.yml
#2154 opened Aug 27, 2024 by inkimikoko Loading…
1
fix wrong buffer for oneShotAllReduceKernel under PUSH_MODE
#2099 opened Aug 8, 2024 by YconquestY Loading…
decoder MMHA kernel support INT8 SCALE_Q_INSTEAD_OF_K and SCALE_P_INS…
#2085 opened Aug 5, 2024 by lishicheng1996 Loading…
5
fix GemmFpAIntB MMa::IteratorB::Layout
#2070 opened Jul 31, 2024 by luliyucoordinate Loading…
fix wrong arg in Engine Building Command in docs/source/performance/perf-overview.md Documentation Improvements or additions to documentation
#2057 opened Jul 30, 2024 by RuibaiXu Loading…
4
Fix default min length triaged Issue has been triaged by maintainers
#1935 opened Jul 11, 2024 by akhoroshev Loading…
3
Add support for custom tokenizer and batch size
#1927 opened Jul 9, 2024 by uppalutkarsh Loading…
Dev sm87 trt101
#1880 opened Jul 3, 2024 by sunnyqgg Loading…
Bump transformers from 4.36.2 to 4.38.0 in /examples/multimodal bug Something isn't working dependencies Pull requests that update a dependency file triaged Issue has been triaged by maintainers waiting for feedback
#1689 opened May 28, 2024 by dependabot bot Loading…
@byshiue
2
ProTip! no:milestone will show everything without a milestone.