Skip to content

Actions: lablup/mlxcel

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
perf(qwen35): byte-identical CUDA DFlash verify via per-row RoPE CI #3931: Pull request #2207 synchronize by inureyes
Queued perf/issue-2191-qwen35-dflash-rope
perf/issue-2191-qwen35-dflash-rope
Queued
perf(qwen35): byte-identical CUDA DFlash verify via per-row RoPE CI #3930: Pull request #2207 synchronize by inureyes
2m 7s perf/issue-2191-qwen35-dflash-rope
perf/issue-2191-qwen35-dflash-rope
2m 7s
fix(test): derive the props batch-alias default from the chunk policy… CI #3929: Commit 454500b pushed by inureyes
Queued main
main
Queued
fix(test): derive the props batch-alias default from the chunk policy CI #3928: Pull request #2212 opened by inureyes
Queued fix/issue-2211-props-batch-alias-default
fix/issue-2211-props-batch-alias-default
Queued
fix(rocm): make hipBLASLt GEMM state safe across threads and streams … CI #3927: Commit 3d609e2 pushed by inureyes
7m 55s main
main
7m 55s
fix(rocm): make hipBLASLt GEMM state safe across threads and streams CI #3926: Pull request #2208 synchronize by inureyes
In progress fix/issue-2200-hipblaslt-thread-safety
fix/issue-2200-hipblaslt-thread-safety
In progress
refactor: move KV attention dispatch into the cache (#2171 Phase 4a) CI #3925: Pull request #2210 synchronize by inureyes
23m 26s refactor/issue-2171-kv-attend
refactor/issue-2171-kv-attend
23m 26s
fix(gemma4): make batched MTP verify decode-exact on row-wise geometr… CI #3924: Commit a20ac93 pushed by inureyes
43m 25s main
main
43m 25s
refactor: move KV attention dispatch into the cache (#2171 Phase 4a) CI #3923: Pull request #2210 synchronize by inureyes
18m 3s refactor/issue-2171-kv-attend
refactor/issue-2171-kv-attend
18m 3s
refactor: move KV attention dispatch into the cache (#2171 Phase 4a) CI #3922: Pull request #2210 synchronize by inureyes
1m 46s refactor/issue-2171-kv-attend
refactor/issue-2171-kv-attend
1m 46s
fix(rocm): make hipBLASLt GEMM state safe across threads and streams CI #3921: Pull request #2208 synchronize by inureyes
34m 38s fix/issue-2200-hipblaslt-thread-safety
fix/issue-2200-hipblaslt-thread-safety
34m 38s
refactor: one per-row sampling step and token-bias composition (#2204) CI #3920: Commit 5976af8 pushed by inureyes
21m 58s main
main
21m 58s
refactor: move KV attention dispatch into the cache (#2171 Phase 4a) CI #3919: Pull request #2210 synchronize by inureyes
10m 30s refactor/issue-2171-kv-attend
refactor/issue-2171-kv-attend
10m 30s
fix(gemma4): make batched MTP verify decode-exact on row-wise geometries CI #3918: Pull request #2203 synchronize by inureyes
21m 15s fix/issue-2190-gemma4-batched-mtp-verify
fix/issue-2190-gemma4-batched-mtp-verify
21m 15s
refactor: move KV attention dispatch into the cache (#2171 Phase 4a) CI #3917: Pull request #2210 synchronize by inureyes
21m 56s refactor/issue-2171-kv-attend
refactor/issue-2171-kv-attend
21m 56s
refactor: move KV attention dispatch into the cache (#2171 Phase 4a) CI #3916: Pull request #2210 synchronize by inureyes
3m 11s refactor/issue-2171-kv-attend
refactor/issue-2171-kv-attend
3m 11s
refactor: move KV attention dispatch into the cache (#2171 Phase 4a) CI #3915: Pull request #2210 opened by inureyes
1m 49s refactor/issue-2171-kv-attend
refactor/issue-2171-kv-attend
1m 49s
refactor: one per-row sampling step and token-bias composition CI #3914: Pull request #2204 synchronize by inureyes
19m 52s refactor/issue-2169-row-sampler
refactor/issue-2169-row-sampler
19m 52s
refactor: one prefill plan and one trim hook for the CLI and the serv… CI #3913: Commit d54b209 pushed by inureyes
12m 9s main
main
12m 9s
fix(ssm): preserve float32 timestep computation with BF16 bias CI #3912: Pull request #2209 opened by Jaeyeong-CHOI
Action required Jaeyeong-CHOI:fix/2142-ssm-bf16-timestep
Jaeyeong-CHOI:fix/2142-ssm-bf16-timestep
Action required
fix(rocm): make hipBLASLt GEMM state safe across threads and streams CI #3911: Pull request #2208 opened by inureyes
34m 29s fix/issue-2200-hipblaslt-thread-safety
fix/issue-2200-hipblaslt-thread-safety
34m 29s
refactor: one per-row sampling step and token-bias composition CI #3910: Pull request #2204 synchronize by inureyes
39m 32s refactor/issue-2169-row-sampler
refactor/issue-2169-row-sampler
39m 32s
refactor: one prefill plan and one trim hook for the CLI and the server CI #3909: Pull request #2205 synchronize by inureyes
30m 53s refactor/issue-2170-prefill-plan
refactor/issue-2170-prefill-plan
30m 53s
refactor: one prefill plan and one trim hook for the CLI and the server CI #3908: Pull request #2205 synchronize by inureyes
2m 59s refactor/issue-2170-prefill-plan
refactor/issue-2170-prefill-plan
2m 59s
perf(qwen35): byte-identical CUDA DFlash verify via per-row RoPE CI #3907: Pull request #2207 synchronize by inureyes
45m 32s perf/issue-2191-qwen35-dflash-rope
perf/issue-2191-qwen35-dflash-rope
45m 32s