Skip to content

Pull requests: huggingface/candle

New pull request New
83 Open 1,791 Closed
83 Open 1,791 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

Fix Batcher iterator break condition
#2655 opened Dec 1, 2024 by hhllhhyyds Loading…
Add Nvembed v2 model
#2649 opened Nov 29, 2024 by cdoko Loading…
3
Optimize WCond and ReduceIndex implementations for improved performance and safety
#2648 opened Nov 29, 2024 by pluveto Loading…
Refactor/ops step1 : Introduce OpsRegistry to simplify operators development and enable incremental refactoring (#2638)
#2644 opened Nov 27, 2024 by imihalcea Loading…
3
Sync with GGML: add GGML bf16 support
#2640 opened Nov 26, 2024 by EricLBuehler Loading…
*Major T/s improvement* Use the Metal qmatmul MM kernels
#2615 opened Nov 14, 2024 by EricLBuehler Loading…
1
Add LRSchedulers
#2603 opened Nov 7, 2024 by DimitriTimoz Loading…
Implement BertPooling
#2601 opened Nov 6, 2024 by DimitriTimoz Loading…
fix: usage of a deprecated Node.js version in CI
#2565 opened Oct 15, 2024 by hamirmahal Loading…
1
2
Make Binary cross entropy with logit numerically stable for high logit values
#2562 opened Oct 14, 2024 by BeneSim Loading…
1
Update README.md
#2544 opened Oct 6, 2024 by hackit-coder Loading…
2
Fix result dimension of GRU and implement RNN multi-layer and bidirection
#2542 opened Oct 5, 2024 by kigichang Loading…
Experimentations around cuda graphs
#2538 opened Oct 3, 2024 by LaurentMazare • Draft
[AArch64] Quantized MatMul performance improvement
#2536 opened Oct 3, 2024 by bgergely0 Loading…
Add ability to use any CLIPModel with model-id and revision
#2527 opened Oct 1, 2024 by akshayballal95 Loading…
Clamping t5 hidden states to avoid F16 NaNs
#2481 opened Sep 17, 2024 by EricLBuehler Loading…
1
1
Improve candle_core::Error to make it more ergonomic
#2467 opened Sep 10, 2024 by EricLBuehler Loading…
1
Add the Phi 3.5 MoE model
#2460 opened Sep 3, 2024 by EricLBuehler Loading…
Fix benchmarks on CUDA
#2422 opened Aug 17, 2024 by EricLBuehler Loading…
style: simplify string formatting for readability
#2416 opened Aug 14, 2024 by hamirmahal Loading…
2
Optimize quantization process with QTensor::quantize_onto
#2408 opened Aug 10, 2024 by EricLBuehler Loading…
Add Tensor::unfold, ops::topk_last_dim
#2375 opened Aug 1, 2024 by EricLBuehler Loading…
Add dp4a for CC < 610
#2372 opened Jul 31, 2024 by EricLBuehler Loading…
Add masked_fill under Tensor
#2346 opened Jul 20, 2024 by fgsch Loading…
2
Refactor: Unify f64-Tensor arithmetic operations using macro
#2314 opened Jul 7, 2024 by monadbobo Loading…
ProTip! What’s not been updated in a month: updated:<2024-11-02.