-
Notifications
You must be signed in to change notification settings - Fork 53
Pull requests: mudler/vllm.cpp
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
feat(BACKEND-ROCM): register kFusedNormRope, so GLM-5.3 stops taking the path a keep-quant MLA checkpoint cannot survive
#2572
opened Sep 2, 2026 by
localai-org-maint-bot
Collaborator
Loading…
record(BACKEND-TENSTORRENT-HOST-FREE-FORWARD): the captured-arm verdict is token-clean at three sizes (#2566)
#2567
opened Sep 2, 2026 by
lu-zero
Collaborator
Loading…
fix(FIX-UNALIGNED-LOADERS-772): refuse to borrow safetensors bytes that sit at an odd element offset (#2558)
#2561
opened Sep 2, 2026 by
lu-zero
Collaborator
Loading…
diag(MODEL-MM-QWEN4-EXP): the MoE expert selections DIFFER, and a third of the boundaries are exact bf16 ties (#2552)
#2556
opened Sep 2, 2026 by
localai-org-maint-bot
Collaborator
Loading…
ci(ENG-CI-ORACLE-GOLDENS): run the oracle-goldens digests on a lane, not only on a developer box
#2555
opened Sep 2, 2026 by
localai-org-maint-bot
Collaborator
Loading…
dots3-note W6a: the DENSE vision tower, on a SERVED image request
#2523
opened Sep 1, 2026 by
localai-org-maint-bot
Collaborator
Loading…
perf(BACKEND-ROCM): cooperate across Q8_K activation blocks
#2472
opened Sep 1, 2026 by
VikashLoomba
Contributor
Loading…
11 tasks done
feat(MODEL-MM-GLM53-FLASH-KPOOL-CUDA): give GLM-5.3-Flash's k-pool indexer a device it can run on
#2432
opened Aug 31, 2026 by
localai-org-maint-bot
Collaborator
Loading…
docs: align GLM-5.3 public status
#2374
opened Aug 31, 2026 by
localai-org-maint-bot
Collaborator
Loading…
perf(ROCm): replace software Dp4a with hardware v_dot4_i32_iu8
#2363
opened Aug 30, 2026 by
ghazni101
Contributor
Loading…
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel
#2361
opened Aug 30, 2026 by
ghazni101
Contributor
Loading…
feat(BACKEND-VULKAN-TQ1_0): TQ1_0 ternary keep-quant matmul, MoE, and rope shaders for Vulkan
#2248
opened Aug 29, 2026 by
phantomic12
Loading…
docs: align FP8 KV and placement claims
#2236
opened Aug 29, 2026 by
localai-org-maint-bot
Collaborator
Loading…
perf(GFX1100-TG200): the 200 tok/s gfx1100 decode campaign and fp8 KV cache decode attention
#1936
opened Aug 25, 2026 by
ghazni101
Contributor
Loading…
ProTip!
Type g i on any issue or pull request to go back to the issue listing page.