-
Notifications
You must be signed in to change notification settings - Fork 1.7k
Pull requests: huggingface/candle
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Initial integration of cutile and fused MoE kernel
#3934
opened Aug 31, 2026 by
EricLBuehler
Member
Loading…
Fix the symphonia-gated examples build against the 0.6 audio API
#3931
opened Aug 31, 2026 by
HuzaifaAbdulRehman
Loading…
chore(deps): update hf-hub requirement from 0.5.0 to 1.0.0
dependencies
Pull requests that update a dependency file
rust
Pull requests that update rust code
#3928
opened Aug 31, 2026 by
dependabot
Bot
Loading…
fix: honor decoupled head_dim in quantized_llama and quantized_qwen2
#3917
opened Aug 27, 2026 by
rexlunae
Loading…
feat(cuda): enable runtime dynamic loading and gate MoE cudart linking
#3916
opened Aug 25, 2026 by
Sadik00789
Loading…
fix(onnx): Tile with a zero repeat must empty the dimension
#3913
opened Aug 24, 2026 by
shoemoney
Loading…
fix(quantized): QTensor::quantize ignores layout offset on a narrowed-then-contiguous view
#3912
opened Aug 23, 2026 by
philphauler
Loading…
fix: align F8E4M3 CUDA kernel names with DType::as_str()
#3911
opened Aug 22, 2026 by
philphauler
Loading…
fix(kernels): gate __hmax_nan/__hmin_nan shim on toolkit version
#3909
opened Aug 20, 2026 by
jorge-menjivar
Contributor
Loading…
Add fused chunked-scan kernels for gated-DeltaNet prefill
#3906
opened Aug 18, 2026 by
alytaphoenix
Loading…
fix(gemma3): make value_states contiguous before the KV cache append
#3902
opened Aug 17, 2026 by
sidd-27
Loading…
Add MXFP4 (GGUF dtype 39) to candle-core's quantized backend
#3899
opened Aug 16, 2026 by
rexlunae
Loading…
Metal QMatMul: extend the batch==1 mat-vec fast path to batch<=8
#3898
opened Aug 15, 2026 by
alytaphoenix
Loading…
Add fused Metal kernels for gated-DeltaNet decode and preprocessing
#3897
opened Aug 15, 2026 by
alytaphoenix
Loading…
4 tasks done
Previous Next
ProTip!
Type g i on any issue or pull request to go back to the issue listing page.