Skip to content

Actions: mudler/vllm.cpp

Actions

ci

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
ci.yml
2,500+ workflow runs
2,500+ workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2554: Pull request #562 synchronize by localai-bot
5m 10s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
5m 10s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2553: Pull request #562 synchronize by localai-bot
5m 56s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
5m 56s
docs(dspark): storage refuted; ratio stable at ~0.966 over three with… ci #2552: Commit 72e661a pushed by localai-bot
46m 42s main
main
46m 42s
MODEL-NEMOTRON-H W3 repair: ClaimMamba ignored quantized, and the legacy-alias precedence was INVERTED (#517) ci #2551: Pull request #572 opened by localai-bot
43m 38s row/MODEL-NEMOTRON-H-W3-FIX
row/MODEL-NEMOTRON-H-W3-FIX
43m 38s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2550: Pull request #562 synchronize by localai-bot
4m 59s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
4m 59s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2549: Pull request #562 synchronize by localai-bot
11m 32s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
11m 32s
L9a: the first-party NVFP4 DiT loads -- it was SWIZZLED and HIGH-nibble-first, not linear (#435) ci #2548: Pull request #571 opened by localai-bot
1h 38m 24s row/LTX25-L9A-NVFP4-LINEAR
row/LTX25-L9A-NVFP4-LINEAR
1h 38m 24s
fix(ROCM): AttnQkNormRopeGate output-dtype dispatch -- the 0.8B GDN forward divergence root-caused and fixed (RED anchor turns GREEN) ci #2547: Pull request #570 synchronize by VikashLoomba
1h 29m 39s VikashLoomba:row/ROCM-GDN-08B-FIX
VikashLoomba:row/ROCM-GDN-08B-FIX
1h 29m 39s
fix(ROCM): AttnQkNormRopeGate output-dtype dispatch -- the 0.8B GDN forward divergence root-caused and fixed (RED anchor turns GREEN) ci #2546: Pull request #570 opened by VikashLoomba
32s VikashLoomba:row/ROCM-GDN-08B-FIX
VikashLoomba:row/ROCM-GDN-08B-FIX
32s
gate(ROCM): M4 for Qwen3.5-0.8B GDN -- the oracle-backed near-tie gate as the RED anchor of the forward-divergence fix ci #2545: Pull request #559 synchronize by VikashLoomba
2h 7m 47s VikashLoomba:row/ROCM-M4-GDN-GATE
VikashLoomba:row/ROCM-M4-GDN-GATE
2h 7m 47s
gate(ROCM): M4 -- pinned vLLM-ROCm oracle on gfx1100 + ROCm device-golden lane in the Qwen3-0.6B near-tie gate ci #2544: Pull request #549 synchronize by VikashLoomba
1h 36m 37s VikashLoomba:row/ROCM-M4-GATE
VikashLoomba:row/ROCM-M4-GATE
1h 36m 37s
fix(ltx-2.5): the audio VAE's norm_eps was read by NO arm, and the keyframe refusal named a reason L11 removed (#560) ci #2543: Pull request #563 synchronize by localai-bot
1h 48m 39s row/LTX25-EPS-PIN
row/LTX25-EPS-PIN
1h 48m 39s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2542: Pull request #562 synchronize by localai-bot
7m 27s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
7m 27s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2541: Pull request #562 synchronize by localai-bot
12m 14s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
12m 14s
spec(BACKEND-ROCM): the ROCm head_dim=128 decode arm, the ROCm half of #382 ci #2540: Pull request #564 synchronize by joral
1h 40m 23s joral:row/ROCM-DECODE-ATTN-D128-SPEC
joral:row/ROCM-DECODE-ATTN-D128-SPEC
1h 40m 23s
LTX-2.5: 21B joint video+audio DiT and the generalized video seam ci #2539: Pull request #437 synchronize by localai-bot
1h 37m 38s row/MODEL-DIFFUSION-LTX25
row/MODEL-DIFFUSION-LTX25
1h 37m 38s
BENCH-ASSERT-CLOCK-STATE: the SM clock moves 12.8% between boots and nobody recorded it (#543, #545) ci #2538: Pull request #568 opened by localai-bot
1h 3m 17s row/BENCH-ASSERT-CLOCK-STATE-FIX
row/BENCH-ASSERT-CLOCK-STATE-FIX
1h 3m 17s
MODEL-NEMOTRON-H W1 land-prep: the MIXED_PRECISION resolver re-merged… ci #2537: Commit 1bc5ef8 pushed by localai-bot
49m 51s main
main
49m 51s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2536: Pull request #562 synchronize by localai-bot
29m 48s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
29m 48s
W2 finish (KERNEL-SSM-MAMBA): 9/9 CUDA mutations caught, compute-sanitizer clean, Debug arm green, spec 8.3 re-authored (#496) ci #2535: Pull request #566 opened by localai-bot
1h 34m 26s row/KERNEL-SSM-MAMBA-SSD-W2-FINISH
row/KERNEL-SSM-MAMBA-SSD-W2-FINISH
1h 34m 26s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2534: Pull request #562 synchronize by localai-bot
20m 55s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
20m 55s
MODEL-NEMOTRON-H W3: NemotronH is KNOWN -- registered, parsed, enumerated, KV-shaped (#517) ci #2533: Pull request #565 opened by localai-bot
1h 37m 26s row/MODEL-NEMOTRON-H-W3C
row/MODEL-NEMOTRON-H-W3C
1h 37m 26s
spec(BACKEND-ROCM): the ROCm head_dim=128 decode arm, the ROCm half of #382 ci #2532: Pull request #564 opened by joral
49m 11s joral:row/ROCM-DECODE-ATTN-D128-SPEC
joral:row/ROCM-DECODE-ATTN-D128-SPEC
49m 11s
fix(ltx-2.5): the audio VAE's norm_eps was read by NO arm, and the keyframe refusal named a reason L11 removed (#560) ci #2531: Pull request #563 opened by localai-bot
1h 19m 13s row/LTX25-EPS-PIN
row/LTX25-EPS-PIN
1h 19m 13s
perf(dspark): unblock upstream ncu; the DRAM attribution is CORROBORATED (#442) ci #2530: Pull request #562 opened by localai-bot
17m 36s row/SPEC-DSPARK-MARLIN-NCU
row/SPEC-DSPARK-MARLIN-NCU
17m 36s