rblundon Ryan Blundon
0 Followers · 0 Following
Joined on 2026-05-07
Repositories
1
 Projects Packages Public Activity Starred Repositories
rblundon pushed to main at rblundon/homelab 2026-08-16 22:39:46 -05:00
a2994bf55d feat(astro-orbiter): bump Qwen3.8-27B ctx-size 32768->131072 (128K) [t_441470b9]
rblundon pushed to main at rblundon/homelab 2026-08-16 20:41:01 -05:00
7b44a41da3 feat(llm): swap astro-orbiter primary model Qwen3.6 -> Qwen3.8-27B-Q4_K_M
rblundon pushed to main at rblundon/homelab 2026-08-15 00:13:02 -05:00
efaff340a4 openviking: fix VLM model alias and lower max_input_tokens to 1024
rblundon pushed to main at rblundon/homelab 2026-08-14 23:22:41 -05:00
48536f2615 fix(openviking): cap embedding max_input_tokens at 1536 to stay under llama.cpp nomic-bert 2048 ctx limit
rblundon pushed to main at rblundon/homelab 2026-08-14 12:49:54 -05:00
170a31d090 feat(openviking): deploy maelstrom-ui Web Studio frontend
rblundon pushed to main at rblundon/homelab 2026-08-14 00:16:29 -05:00
aa2730efd5 fix: OpenViking ingress TLS issuer from letsencrypt-internal to letsencrypt-prod
rblundon pushed to main at rblundon/homelab 2026-08-13 23:49:36 -05:00
0dbb77b023 Fix OpenViking: 1Password item mismatch + invalid embedding config fields
rblundon pushed to main at rblundon/homelab 2026-08-13 23:43:13 -05:00
fee9965d0a fix: OpenViking sync-wave deadlock - move ExternalSecret ordering inside Application
rblundon pushed to main at rblundon/homelab 2026-08-13 23:37:14 -05:00
d0f3ddba0d OpenViking application.yaml: fix invalid Helm chart source (chart -> path)
rblundon pushed to main at rblundon/homelab 2026-08-13 23:33:58 -05:00
d9e41118f8 feat(openviking): pilot deployment to fastpass (wave 8)
rblundon pushed to main at rblundon/homelab 2026-08-13 23:18:43 -05:00
ad70b3439c feat(llm): add nomic-embed-text-v1.5-Q4_K_M to astro-orbiter router
rblundon pushed to main at rblundon/homelab 2026-08-13 18:47:18 -05:00
a04435ee9b fix(monitoring): drop Qwen3.6 scrape job — causes CUDA OOM on each scrape (t_02c15dae)
rblundon pushed to main at rblundon/homelab 2026-08-13 15:44:13 -05:00
a2ddb65425 fix(monitoring): reduce llama-server scrape_interval 15s -> 90s to allow GPU P8 idle
rblundon pushed to main at rblundon/homelab 2026-08-13 09:08:22 -05:00
a87da82ebd fix: correct astro-orbiter llama-server scrape target for router cutover
rblundon pushed to main at rblundon/homelab 2026-08-13 09:07:30 -05:00
7aea88724f Add Qwen2.5-Coder-14B-Instruct-4bit to astro-orbiter router (t_55c164f5)
rblundon pushed to main at rblundon/homelab 2026-08-12 23:33:44 -05:00
6455d22752 feat(llm-router): add Meta-Llama-3.1-8B-Instruct-4bit alias; document Phi json_schema limitation (t_9adf0889)
rblundon pushed to main at rblundon/homelab 2026-08-12 22:58:43 -05:00
a47b29d49f feat(llm-router): switch to --models-preset mode; add Phi-3.5-mini-instruct-8bit alias (t_9adf0889)
rblundon pushed to main at rblundon/homelab 2026-08-12 22:27:19 -05:00
9c969f783d feat(llm-inference-multimodel): bump router --models-max 1->4 on astro-orbiter (t_33acbb2e)
rblundon pushed to main at rblundon/homelab 2026-08-12 22:20:02 -05:00
081156ecab feat(llm-inference-multimodel): codify Phi-3.5-mini + Llama-3.1-8B GGUF staging (t_730f9584)
3783ded62a fix: update router unit template comment — no longer a shadow deployment (t_cd0d5388)
5a2246a540 feat: add day2_cutover_qwen_to_router.yml playbook (t_cd0d5388)
ba311a3ec6 feat(llm-inference): add llama.cpp router mode shadow deployment
d1f97ad5ac Phase 2 revised: consolidate astro-orbiter to single Qwen2.5-14B-1M model (port 8002)
Compare 5 commits »
rblundon pushed to main at rblundon/homelab 2026-08-06 10:40:21 -05:00
b4bdb63e4a llm-inference-multimodel: correct stale VRAM estimate for qwen-1m shadow slot