Skip to content

Actions: gittensor-ai-lab/sparkinfer

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
2,500+ workflow runs
2,500+ workflow runs

Filter by Workflow

⌥ + click/return to exclude

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
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default rtx5090-required #307: Pull request #601 edited by James-CUDA
6s
6s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default build-attested-binaries #206: Pull request #601 synchronize by James-CUDA
11m 31s James-CUDA:perf/qwen36-moe-next-opt
James-CUDA:perf/qwen36-moe-next-opt
11m 31s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default sensitive-paths-guard #986: Pull request #601 synchronize by James-CUDA
10s
10s
fix(qwen36): restore MoE device-tilemap (MOE_GPU) correctness — masked down-dequant no-op cap-open-prs #415: Pull request #602 opened by James-CUDA
6s
6s
fix(qwen36): restore MoE device-tilemap (MOE_GPU) correctness — masked down-dequant no-op build-attested-binaries #205: Pull request #602 opened by James-CUDA
9m 58s James-CUDA:fix/qwen36-moe-gpu-mask-dequant
James-CUDA:fix/qwen36-moe-gpu-mask-dequant
9m 58s
fix(qwen36): restore MoE device-tilemap (MOE_GPU) correctness — masked down-dequant no-op rtx5090-required #306: Pull request #602 opened by James-CUDA
8s
8s
fix(qwen36): restore MoE device-tilemap (MOE_GPU) correctness — masked down-dequant no-op copycat-guard #334: Pull request #602 opened by James-CUDA
15s
15s
fix(qwen36): restore MoE device-tilemap (MOE_GPU) correctness — masked down-dequant no-op sensitive-paths-guard #985: Pull request #602 opened by James-CUDA
6s
6s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default build-attested-binaries #204: Pull request #601 synchronize by James-CUDA
9m 19s James-CUDA:perf/qwen36-moe-next-opt
James-CUDA:perf/qwen36-moe-next-opt
9m 19s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default sensitive-paths-guard #984: Pull request #601 synchronize by James-CUDA
5s
5s
Merge pull request #598 from inference2026/feat/qwen36-512-prefill build-attested-binaries #203: Commit 86cfb59 pushed by skyrocket2026
11m 24s main
main
11m 24s
pages build and deployment pages-build-deployment #657: by skyrocket2026
44s main
main
44s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default build-attested-binaries #202: Pull request #601 opened by James-CUDA
10m 12s James-CUDA:perf/qwen36-moe-next-opt
James-CUDA:perf/qwen36-moe-next-opt
10m 12s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default copycat-guard #333: Pull request #601 opened by James-CUDA
13s
13s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default cap-open-prs #414: Pull request #601 opened by James-CUDA
8s
8s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default rtx5090-required #305: Pull request #601 opened by James-CUDA
7s
7s
qwen36 MoE prefill: fix MOE_GPU device-tilemap correctness (#586) + int8 projection default sensitive-paths-guard #983: Pull request #601 opened by James-CUDA
6s
6s
perf(qwen36): default MoE live-expert gather on (+5.3% pp @512) rtx5090-required #304: Pull request #598 edited by inference2026
9s
9s
perf(qwen36): default MoE live-expert gather on (+5.3% pp @512) build-attested-binaries #201: Pull request #598 synchronize by inference2026
9m 22s inference2026:feat/qwen36-512-prefill
inference2026:feat/qwen36-512-prefill
9m 22s
perf(qwen36): default MoE live-expert gather on (+5.3% pp @512) sensitive-paths-guard #982: Pull request #598 synchronize by inference2026
8s
8s
Merge pull request #595 from fansilas/fix/qwen36-restore-moe-fp8-default build-attested-binaries #200: Commit bb132f3 pushed by skyrocket2026
10m 27s main
main
10m 27s
pages build and deployment pages-build-deployment #656: by skyrocket2026
37s main
main
37s
Merge pull request #600 from gittensor-ai-lab/fix/cb-ttft-no-regressi… eval-policy #276: Commit f02462b pushed by skyrocket2026
10s main
main
10s
pages build and deployment pages-build-deployment #655: by skyrocket2026
36s main
main
36s
fix(eval): reject CB mixed TTFT regressions and harden harness eval-policy #275: Pull request #600 opened by skyrocket2026
9s fix/cb-ttft-no-regression-gate
fix/cb-ttft-no-regression-gate
9s