Frank 048e7eaaba frank-default-large: full 262144 context, 1 slot ...

User request 2026-08-19 (second pass): parallel slots degraded
performance. PARALLEL=1, CTX_SIZE=262144 (native max, -c equals the
per-slot context with one slot). Concurrent requests queue on the
single slot; the server never preempts a running generation.
pi models.json contextWindow/maxTokens back to 262144 (outside repo).
Docs updated to match.

2026-08-19 11:17:31 -07:00

..

deploy.sh

Swap fleet naming + restore Qwen3.6-35B-A3B Q6_K_P on Dankserver

2026-08-18 00:32:26 -07:00

frank-default-large.env

frank-default-large: full 262144 context, 1 slot

2026-08-19 11:17:31 -07:00

frank-default-large.service

Swap fleet naming + restore Qwen3.6-35B-A3B Q6_K_P on Dankserver

2026-08-18 00:32:26 -07:00

frank-default-small.env

Swap fleet naming + restore Qwen3.6-35B-A3B Q6_K_P on Dankserver

2026-08-18 00:32:26 -07:00

frank-default-small.service

Swap fleet naming + restore Qwen3.6-35B-A3B Q6_K_P on Dankserver

2026-08-18 00:32:26 -07:00

serve-frank-large.sh

Reduce ctxcp from 32 to 16 to free ~16 GiB VRAM headroom and prevent OOM crashes during prompt processing

2026-08-18 16:11:58 -07:00

serve-frank-small.sh

Swap fleet naming + restore Qwen3.6-35B-A3B Q6_K_P on Dankserver

2026-08-18 00:32:26 -07:00