Skip to content

Actions: kmbandy/llama.cpp

Actions

Server

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
server.yml
145 workflow runs
145 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
wp-worker/vulkan: gate all persistent-graphs behavior behind the flag… Server #120: Commit 9e52f53 pushed by kmbandy
13m 4s master
master
13m 4s
vulkan: fix rvalue binding of marked subbuffers in mul_mat_q_f16 Server #119: Commit 9ae8ff8 pushed by kmbandy
16m 40s master
master
16m 40s
vulkan: persistent graph plans (WP_PERSISTENT_GRAPHS); mtp: serialize… Server #118: Commit 965a65c pushed by kmbandy
12m 50s master
master
12m 50s
pipeline: bit-exact wire compression for LAN frames (WP_WIRE_COMPRESS) Server #117: Commit 105cfe0 pushed by kmbandy
15m 19s master
master
15m 19s
vulkan: enqueue cached host-read staging copy before ctx_end Server #116: Commit ddaf823 pushed by kmbandy
12m 30s master
master
12m 30s
wp-worker: record wire stats in multi-device mode; vulkan: cached hos… Server #115: Commit c254acd pushed by kmbandy
13m 31s master
master
13m 31s
wp-worker: vulkan-safe WP_SUBMIT_ASYNC + per-size-class pin caps Server #114: Commit ec11851 pushed by kmbandy
13m 15s master
master
13m 15s
pipeline/worker: per-socket coalesce modes and recv/send instrumentation Server #113: Commit b158cfc pushed by kmbandy
13m 22s master
master
13m 22s
wp-expert-worker: WP_EXPERT_COUNTS_DUMP — persist learned expert hotness Server #112: Commit 27b16a7 pushed by kmbandy
12m 36s master
master
12m 36s
ggml-cuda/qwen4exp: kill redundant graph-key host passes, unblock hc … Server #111: Commit 3012cab pushed by kmbandy
12m 24s master
master
12m 24s
ggml-cpu: WP_CPU_GEMM_SCALE_BCAST — broadcast q4_K scales without vps… Server #110: Commit 83254fa pushed by kmbandy
12m 40s master
master
12m 40s
docs: regenerate the GitNexus agent-instruction block Server #109: Commit 3ec6a7a pushed by kmbandy
13m 54s master
master
13m 54s
qwen4exp: sync with upstream's merged Qwen3.8-Flash-Next (ggml-org#27… Server #108: Commit 3e0ec4d pushed by kmbandy
16m 13s master
master
16m 13s
wp-expert-worker: per-class arenas, so multi-size-class plans can be … Server #107: Commit 9e49eb0 pushed by kmbandy
15m 0s master
master
15m 0s
wp-expert-descriptor: recognize qwen4exp as SwiGLU Server #106: Commit 4d75c19 pushed by kmbandy
18m 20s master
master
18m 20s
Merge branch 'ds4/policy': expert LFU history and pinning Server #105: Commit d717f0b pushed by kmbandy
12m 25s master
master
12m 25s
llama-context: cap the logits buffer only where the graph caps the LM… Server #104: Commit df09d8d pushed by kmbandy
13m 42s master
master
13m 42s
llama-arch: deny -sm tensor for qwen4exp Server #103: Commit 1cecef9 pushed by kmbandy
15m 24s master
master
15m 24s
docs/dev: 08-08/08-09 ledgers, eviction sim policies, section-4 falsi… Server #102: Commit 51c70e0 pushed by kmbandy
16m 23s master
master
16m 23s
fix(cuda): solve_tri k-tiling must not early-return before the sA bar… Server #101: Commit ae1330a pushed by kmbandy
11m 12s master
master
11m 12s
merge: link-order fix from 2026 with the worker resource fixes from main Server #100: Commit 0f1034a pushed by kmbandy
14m 55s master
master
14m 55s
build: fix expert-worker link order for the GPU builds Server #99: Commit 66d6bb7 pushed by kmbandy
15m 56s master
master
15m 56s
wp-expert-shard: emit sidecars from --manifest-only Server #98: Commit f17a5a0 pushed by kmbandy
15m 4s master
master
15m 4s
merge: design-doc update from 2026 with the dense extractor from main Server #97: Commit 8fc301f pushed by kmbandy
15m 29s master
master
15m 29s
build: factor the expert-dispatch sources into a static library Server #96: Commit c553d6a pushed by kmbandy
18m 37s master
master
18m 37s