-
Notifications
You must be signed in to change notification settings - Fork 33
Pull requests: mudler/vllm.cpp
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
record(dspark): the oracle varies 20% between invocations — parity is UNMEASURED (#442)
#855
opened Aug 15, 2026 by
localai-bot
Collaborator
Loading…
docs: align speech generation surfaces
#854
opened Aug 15, 2026 by
localai-org-maint-bot
Contributor
Loading…
feat(ENG-RECORD-ANCHOR-RATCHET): parse the citation form the checker never read, and ratchet its rot
#851
opened Aug 15, 2026 by
localai-bot
Collaborator
Loading…
fix(GATE-SQUASH-TRAILERS): the PR body is the commit message now, so gate it like one
#850
opened Aug 15, 2026 by
localai-bot
Collaborator
Loading…
2
feat(VT-FP8-W8A8-CPU-ARM): register the static fp8 W8A8 path on the CPU backend (#468)
#842
opened Aug 14, 2026 by
localai-bot
Collaborator
Loading…
4 tasks done
porting-inventory: a smaller Qwen3.8 sibling DOES exist, and it loads (#740)
#823
opened Aug 14, 2026 by
localai-bot
Collaborator
Loading…
fix(capi): hoist SpeechRegistry() out of extern "C" -- main is RED on clang/hipcc (#805)
#814
opened Aug 14, 2026 by
VikashLoomba
Contributor
Loading…
perf(rocm): head_dim=128 decode arm -- the ROCm half of #382, default OFF
#767
opened Aug 14, 2026 by
joral
Contributor
Loading…
MODEL-NEMOTRON-H: the WEIGHT LOADER — the real checkpoint LOADS in its packed format (the model is NOT yet reachable from the ABI, see #810)
#752
opened Aug 14, 2026 by
localai-bot
Collaborator
Loading…
fix(LTX25-TILED-DECODE): chunking binds at 81 frames, and the untiled-axis mapper had no gate (#644)
#746
opened Aug 14, 2026 by
localai-bot
Collaborator
Loading…
docs: align registry and speech status
#696
opened Aug 14, 2026 by
localai-org-maint-bot
Contributor
Loading…
feat(tenstorrent): host-free decode forward — capture investigation + architecture
#694
opened Aug 13, 2026 by
lu-zero
Contributor
Loading…
spec(LTX25-KEYFRAMES-ABS-POS): the module that refuses BOTH shipped checkpoints (#658)
#678
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
record(GEMMA4-ROCM-KEEP): measured SharedK-WMMA plateau on 2x R9700
#676
opened Aug 13, 2026 by
bakon11
Contributor
Loading…
fix(LTX25-DEVICE-SEAM-SIBLING): the third question, and a leakage bucket that is a property rather than a spelling (#659, #660)
#671
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
record(LTX25-PROMPT-ADALN): the prompt-AdaLN defect was understated ~138x, and the keyframes claim next door was false (#644)
#663
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
feat(ltx-2.5): port the temporal x2 latent upsampler, and say what drives it (#644)
#662
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
fix(LTX25-RETIRE-DEAD-ARMS): retire the invented refusals, refuse the ignored extra (#644, #611)
#661
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
feat(LTX25-TILED-DECODE): tiled + STREAMING Conv VAE decode, and the 60 GiB is not in the decode (#644)
#656
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
tools(GATE-GPU-LOCK-WRAPPER): one sanctioned way to take the GPU, and it records which lock it took (#587)
#596
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
fix(tests): the Windows arm cannot compile test_backend_cross_device (#514, #540)
#578
opened Aug 13, 2026 by
localai-bot
Collaborator
Loading…
2
fix(ROCM): AttnQkNormRopeGate output-dtype dispatch -- the 0.8B GDN forward divergence root-caused and fixed (RED anchor turns GREEN)
#570
opened Aug 13, 2026 by
VikashLoomba
Contributor
•
Draft
gate(ROCM): M4 for Qwen3.5-0.8B GDN -- the oracle-backed near-tie gate as the RED anchor of the forward-divergence fix
#559
opened Aug 13, 2026 by
VikashLoomba
Contributor
•
Draft
gate(ROCM): M4 -- pinned vLLM-ROCm oracle on gfx1100 + ROCm device-golden lane in the Qwen3-0.6B near-tie gate
#549
opened Aug 12, 2026 by
VikashLoomba
Contributor
•
Draft
Previous Next
ProTip!
Filter pull requests by the default branch with base:main.