phase 9: KV cache + autoregressive generation

- KVCache: per-layer, per-head storage with append + reconstruct
- forward_with_cache: prefill (full prompt) + decode (single token) modes
- Fixed data layout bug: per-head vectors avoid cross-head interleaving
- CLI updated to use KV cache by default
- bench-gpt2 supports --no-cache flag for comparison

Benchmark results (50 prompts × 20 tokens):
- KV cache vs no-cache: 50/50 bit-identical (cache is correct)
- 18x speedup: TTFT 400→24ms, TBT 407→22ms, throughput 2.5→44 tok/s
- vs HF transformers: 40/50 match (10 are FP divergence, avg logit gap 0.20)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

crates/xserv-model/src/lib.rs

@@ -3,4 +3,4 @@ pub mod gpt2;
 pub mod loader;
 
 pub use config::ModelConfig;
-pub use gpt2::GPT2;
+pub use gpt2::{GPT2, KVCache};