Restore historical Q6 large model runtime

scripts/frank-default-large.env

@@ -9,8 +9,11 @@
 # HIP runtime reads those bare env vars (see the 2026-08-13 TEMP incident).
 # llama.cpp reads --host/--port/--temp as CLI flags, never as env vars.
 
-# ROCmFPX build with TurboQuant KV-cache support.
+# PromptForge ROCmFPX build and its historical ROCm 7.15 SDK environment.
 LLAMA_BIN=/home/dank/ciru-qwen38/ROCmFPX/build-promptforge/bin/llama-server
+ROCM_DEVEL=/home/dank/ling3-ciru/Ling-3.0-Flash-CIRU-int4-Strix-native/.venv/lib/python3.12/site-packages/_rocm_sdk_devel
+ROCM_LIB=/home/dank/ling3-ciru/Ling-3.0-Flash-CIRU-int4-Strix-native/.venv/lib/python3.12/site-packages/_rocm_sdk_libraries
+ROCM_CORE=/home/dank/ling3-ciru/Ling-3.0-Flash-CIRU-int4-Strix-native/.venv/lib/python3.12/site-packages/_rocm_sdk_core
 
 # Model files from https://huggingface.co/kingjones777/Qwen3.8-27B-Uncensored-ROCmFP4-STRIX-MTP-GGUF.
 MODEL_DIR=/home/dank/models/kingjones-Qwen3.8-27B-STRIX-MTP
@@ -37,13 +40,13 @@ LISTEN_PORT=7770
 CTX_SIZE=262144
 PARALLEL=1
 # Batch/ubatch follow the existing qualified Dankclaw profile.
-BATCH_SIZE=512
-UBATCH_SIZE=512
+BATCH_SIZE=4096
+UBATCH_SIZE=4096
 NUM_THREADS=16
 THREAD_BATCH=32
 CACHE_TYPE_K=f16
-CACHE_TYPE_V=turbo4
-CTX_CHECKPOINTS=32
+CACHE_TYPE_V=f16
+CTX_CHECKPOINTS=16
 CACHE_RAM=8192
 
 # Standalone MTP draft head, loaded from MTP_DRAFT_FILE by the serve script.