Use ngram-mod and prioritize attention experts

docs/LLM/deployment.md

@@ -125,16 +125,15 @@ The active service on Dankserver is upstream llama.cpp (b10121) serving
 `Ornith-1.5-35B-A3B-CRACK-Q6_K.gguf` under the alias `frank-default-small`
 on port 7770. It runs the stock `/home/dank/llama.cpp/build/bin/llama-server`
 binary with expert tensors selected by an explicit `--override-tensor`
-regex. Expert tensors from full-attention blocks 27 and 31 and the
-output-adjacent blocks 33 through 40 stay on CUDA; other expert tensors are
-CPU-resident. The service uses 7 CPU threads, native `draft-mtp` speculative
-decoding (under re-evaluation after an 8K no-spec benchmark), f16 KV, context
-491520 (3 slots of 163840 each), batch
-4096, ubatch 2048, and `--no-mmap`. The unit is
+regex. Expert tensors from every full-attention block and blocks 35 through 40
+stay on CUDA; other expert tensors are CPU-resident. The service uses 7 CPU
+threads, ngram-mod speculative decoding only, f16 KV, context 491520 (3 slots
+of 163840 each), batch
+4096, ubatch 1024, and `--no-mmap`. The unit is
 `scripts/frank-default-small.service`. It launches through
 `scripts/serve-frank-small.sh`, which reads tunables from
-`scripts/frank-default-small.env`. The chat template comes from
-`peculiar-ragdoll/Qwen-Sharp-Chat-Templates`. See [[dealign-ornith-crack]].
+`scripts/frank-default-small.env`. It uses `ubatch=1024` and the chat template
+from `peculiar-ragdoll/Qwen-Sharp-Chat-Templates`. See [[dealign-ornith-crack]].
 
 ### Dankclaw — frank-default-large