fix(llm): override-kv context_length=65536 so slot isn't capped to 32k

llama-server caps the slot to the GGUF training context (32768) and ignores the
YaRN-extended size, leaving per-seq context at 32k. Raise qwen2.context_length
metadata to 65536 so the full window is served per request.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

stacks/llm/docker-compose.yml

@@ -44,6 +44,8 @@ services:
       - "2"
       - "--yarn-orig-ctx"
       - "32768"
+      - "--override-kv"
+      - "qwen2.context_length=int:65536"
       - "--flash-attn"
       - "on"
       - "--cache-type-k"