Skip to content

Actions: asher/mlx-kquant

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
250 workflow runs
250 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
chore(release): 0.4.17 (#87) release #42: Commit 8688ca9 pushed by asher
40m 56s v0.4.17
v0.4.17
40m 56s
chore(release): 0.4.17 (#87) ci #208: Commit 8688ca9 pushed by asher
44m 11s main
main
44m 11s
chore(release): 0.4.17 ci #207: Pull request #87 opened by asher
18m 4s release-0.4.17
release-0.4.17
18m 4s
perf(kernels): faster sparse-attention top-k and scoring, slot-parall… ci #206: Commit ec935fb pushed by asher
3m 20s main
main
3m 20s
perf(kernels): faster sparse-attention top-k and scoring, slot-parallel MoE down gathers ci #205: Pull request #86 opened by asher
24m 10s perf/q4-select
perf/q4-select
24m 10s
fix(gguf): zero-copy read a tensor past the 1-D window limit at the h… ci #204: Commit 3d5170e pushed by asher
16m 30s main
main
16m 30s
fix(gguf): zero copy read a tensor past the 1-D window limit at the head of its file ci #203: Pull request #85 opened by asher
25m 10s fix/gguf-view-stride
fix/gguf-view-stride
25m 10s
chore(release): 0.4.16 (#84) release #41: Commit ed3c85c pushed by asher
43m 41s v0.4.16
v0.4.16
43m 41s
chore(release): 0.4.16 (#84) ci #202: Commit ed3c85c pushed by asher
43m 37s main
main
43m 37s
chore(release): 0.4.16 ci #201: Pull request #84 opened by asher
29m 31s release-0.4.16
release-0.4.16
29m 31s
feat(codec): q2_0 codec, split-codec shared-expert MoE kernels, faste… ci #200: Commit b683300 pushed by asher
5m 49s main
main
5m 49s
feat(codec): q2_0 codec, split-codec shared-expert MoE kernels, faster router and indexer top-k ci #199: Pull request #83 synchronize by asher
19m 2s codec-q2-0
codec-q2-0
19m 2s
feat(codec): q2_0 codec, split-codec shared-expert MoE kernels, faster router and indexer top-k ci #198: Pull request #83 opened by asher
16m 48s codec-q2-0
codec-q2-0
16m 48s
perf(qmm): split-K for 33-128 activation rows on NAX GPUs (#82) ci #197: Commit 0d19ce2 pushed by asher
16m 51s main
main
16m 51s
perf(qmm): split-K for 33-128 activation rows on NAX GPUs ci #196: Pull request #82 opened by asher
17m 46s perf/nax-midm
perf/nax-midm
17m 46s
chore(release): 0.4.15 (#81) release #40: Commit 83c2e7d pushed by asher
1h 2m 21s v0.4.15
v0.4.15
1h 2m 21s
chore(release): 0.4.15 (#81) ci #195: Commit 83c2e7d pushed by asher
27m 2s main
main
27m 2s
chore(release): 0.4.15 ci #194: Pull request #81 opened by asher
14m 11s release-0.4.15
release-0.4.15
14m 11s
perf(decode): q6_k split-K mat-vec, add_rmsnorm_norm and hd512 decode… ci #193: Commit e596571 pushed by asher
13m 46s main
main
13m 46s
perf(decode): q6_k split-K mat-vec, add_rmsnorm_norm and hd512 decode attention on NAX GPUs ci #192: Pull request #80 opened by asher
17m 33s perf/q6k-splitk-qmv
perf/q6k-splitk-qmv
17m 33s
perf(qmm): split-K for 33-128 activation rows on GPUs without tensor … ci #191: Commit 381c3b7 pushed by asher
13m 58s main
main
13m 58s
perf(qmm): split-K for 33-128 activation rows on GPUs without tensor cores ci #190: Pull request #79 opened by asher
12m 46s perf/midm-occupancy
perf/midm-occupancy
12m 46s
chore(release): 0.4.14 (#78) release #39: Commit e0fb1a9 pushed by asher
1h 4m 16s v0.4.14
v0.4.14
1h 4m 16s
chore(release): 0.4.14 (#78) ci #189: Commit e0fb1a9 pushed by asher
29m 10s main
main
29m 10s
chore(release): 0.4.14 ci #188: Pull request #78 opened by asher
14m 50s release-0.4.14
release-0.4.14
14m 50s