Skip to content

Pull requests: OpenNMT/CTranslate2

New pull request New
18 Open 1,061 Closed
18 Open 1,061 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

small improvement in flash attention
#1732 opened Jun 25, 2024 by minhthuc2502 Loading…
Quantzation AWQ GEMM + GEMV
#1727 opened Jun 19, 2024 by minhthuc2502 Loading…
3
[*] improve log message for storage view content
#1718 opened Jun 4, 2024 by tirivo Loading…
4
Use specific CMAKE_CUDA_COMPILER instead of forcing CMake to use the same compiler for both CXX and CUDA
#1710 opened May 25, 2024 by alexlnkp Loading…
1
add rope alibi to encoder
#1687 opened Apr 29, 2024 by vince62s Loading…
17
Mixtral support
#1644 opened Mar 18, 2024 by minhthuc2502 • Draft
2
CANN Backend support
#1606 opened Jan 26, 2024 by 3manifold Loading…
1
3
build: armv7 and android static omp support
#1600 opened Jan 12, 2024 by ebraraktas Loading…
convert opennmt py on the fly and inference with model in memory
#1578 opened Dec 6, 2023 by minhthuc2502 • Draft
Add RoBERTaConfig to TransformerConverter
#1510 opened Oct 10, 2023 by bhavnicksm Loading…
2
Accept variable-length batch prompts for Whisper
#1457 opened Sep 7, 2023 by guillaumekln • Draft
Accept left offsets when applying position encodings
#1374 opened Jul 28, 2023 by guillaumekln • Draft
Accept left offsets in the rotary embeddings layer
#1372 opened Jul 27, 2023 by guillaumekln • Draft
Accept left offsets in the masked softmax operator
#1370 opened Jul 27, 2023 by guillaumekln • Draft
Keep FFN output layer in float32 for T5 models
#1239 opened May 23, 2023 by guillaumekln • Draft
3
Manually destroy cuBLAS and cuDNN handles before threads exit
#1201 opened Apr 28, 2023 by guillaumekln • Draft
Use non blocking CUDA streams
#1177 opened Apr 20, 2023 by guillaumekln • Draft
Directly register finalized model variables
#1058 opened Jan 27, 2023 by guillaumekln • Draft
ProTip! Filter pull requests by the default branch with base:master.