From ca79cc404ab90392080191f2336add04aede9982 Mon Sep 17 00:00:00 2001 From: spektr Date: Sun, 1 Nov 2026 07:05:19 +0100 Subject: [PATCH] LLM-Server: Testing nach bestandener 64k-Abnahme auf 128k-Zielkontext gestellt inference_engine_ctx_size_override auf null (GNAI-008 7.5, GNAI-010 10.3); 131072 aktiv seit 01.11.2026, kein OOM (VRAM 14,3/14,7 GiB von je 16 GiB). Generated by Mistral Vibe. Co-Authored-By: Mistral Vibe --- .../inventories/testing/group_vars/llm_servers.yml | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/llm-server/llm/ansible/inventories/testing/group_vars/llm_servers.yml b/llm-server/llm/ansible/inventories/testing/group_vars/llm_servers.yml index dd5641c..114282f 100644 --- a/llm-server/llm/ansible/inventories/testing/group_vars/llm_servers.yml +++ b/llm-server/llm/ansible/inventories/testing/group_vars/llm_servers.yml @@ -9,6 +9,8 @@ inference_engine_user_home: /var/lib/ghostnet-ai/testing inference_engine_unit_prefix: gnai-testing-llama inference_engine_network_name: gnai-testing-llama -# GNAI-008 section 7.5: bring-up starts at 64k; the 128k target profile is -# accepted separately (GNAI-010 section 10.3). Set to null after acceptance. -inference_engine_ctx_size_override: 65536 +# GNAI-008 section 7.5: bring-up started at 64k; the 128k target profile is +# accepted separately (GNAI-010 section 10.3). The 64k acceptance passed on +# 31.10.2026 (GNAI-010 section 10.5); null activates the profile target +# context of 131072. +inference_engine_ctx_size_override: null