feat(llm): make spec decoding and KV cache type .env toggles

Move all speculative-decoding flags into LLAMA_SPEC_ARGS (blank = off)
and the KV cache quant into LLAMA_KV_CACHE_TYPE, so token-corruption
A/B tests are a single .env edit instead of a compose edit.
Replaces LLAMA_DRAFT_MODEL_FILE and LLAMA_SPEC_N_MAX.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

docker-compose.yml

@@ -25,12 +25,10 @@ services:
       --parallel ${LLAMA_PARALLEL:-2}
       --kv-unified
       --flash-attn on
-      --cache-type-k q8_0
-      --cache-type-v q8_0
+      --cache-type-k ${LLAMA_KV_CACHE_TYPE:-q8_0}
+      --cache-type-v ${LLAMA_KV_CACHE_TYPE:-q8_0}
       --cache-ram ${LLAMA_CACHE_RAM:-24576}
-      --spec-type draft-dflash
-      -md /models/${LLAMA_DRAFT_MODEL_FILE:-Qwen3.8-27B-DFlash2-Q8_0.gguf}
-      --spec-draft-n-max ${LLAMA_SPEC_N_MAX:-7}
+      ${LLAMA_SPEC_ARGS:-}
       --jinja
     # see docs/architecture.md#llama-server (flags), docs/network-access.md#backends (no host port)
     expose: