code wiki / (root) / nx_specdec_gate.nx

nx_specdec_gate.nx

buildroot/runtime/nx_specdec_gate.nx

8414 B209 linesdepth 15pulls 63 transitivereach 0 importersview sourcekind gate/proof
docsdependenciesstructsconstsfunctions

about

nx_specdec_gate.nx -- MEASURED gate for sovereign SPECULATIVE DECODING (prompt-lookup draft + multi-token greedy verify) on the REAL model. KAT pure drafter: match/no-match/most-recent-match EQUIV-1 chunked prefill (ONE m=n forward) == sequential prefill (n m=1 forwards): first 8 greedy tokens identical EQUIV-2 spec_decode_greedy == plain_greedy: full 24-token sequence identical (proves verify rows + cache truncation + offset- causal positions) SPEED same EQUIV-2 runs timed: us_plain vs us_spec + acceptance stats (echo-y prompt = lookup-friendly; honest workload note) license_tier: ORIGINAL expect_exit: 0

dependencies 23 imports · 0 importers

nx_syscalls.nx nx_tier.nx nx_le.nx nx_bpe.nx nx_gguf.nx nx_gguf_load.nx nx_gguf_meta.nx nx_f32.nx nx_f32_kv_cache.nx nx_f32_lazy_weight.nx nx_specdec_gate.nx

diagram shows first 10 each side; +13 more imports, +0 more importers in the complete lists below.

imports: nx_syscalls.nxnx_tier.nxnx_le.nxnx_bpe.nxnx_gguf.nxnx_gguf_load.nxnx_gguf_meta.nxnx_f32.nxnx_f32_kv_cache.nxnx_f32_lazy_weight.nxnx_f32_llama_block.nxnx_f32_llama_block_v4.nxnx_f32_llama_stack_v4.nxnx_f32_llama_layer_lazy_load.nxnx_f32_llm.nxnx_f32_llm_v4.nxnx_f32_llm_read_dims.nxnx_f32_bpe_load.nxnx_f32_llm_special_tokens.nxnx_f32_sampler.nxnx_prng.nxnx_reasoning.nxnx_specdec.nx

imported by: nobody (leaf or entry point)

call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown

main sys_mmap nx_spec_draft_lookup sg_w sys_write sys_read_file sys_openat_rd sys_lseek sys_mmap ↻ sys_read sys_close nx_gguf_parse nx_le_read_u32 nx_le_read_u64 nx_le_read_u32 ↻ _gguf_skip_value nx_le_read_u64 ↻ nx_le_read_u32 ↻ _gguf_skip_value ↻ sys_mmap ↻ nx_f32_llama_model_alloc sys_mmap ↻ nx_f32_llm_read_dims_from_ sys_mmap ↻ nx_gguf_meta_find nx_le_read_u64 ↻ nx_le_read_u32 ↻ _gmeta_bytes_equal nx_gguf_skip_value _gguf_skip_value ↻ nx_gguf_meta_read_string_l nx_le_read_u64 ↻ nx_gguf_meta_read_string_p _fld_read_arch_u32 sys_mmap ↻ _fld_concat_key nx_gguf_meta_find ↻ nx_gguf_meta_read_u32 nx_le_read_u32 ↻ _fld_read_vocab_size_from_

structs

none

consts

none

functions

39func sg_w(s: *u8) -> i64 { var n: i64 = 0; while s[n] != (0 as u8) { n = n + 1 } sys_write(1, s, n); return 0 }
called by 2: sg_wnmain calls 1: sys_write
40func sg_wn(v: i64) -> i64
called by 1: main calls 3: sg_wsys_mmapsys_write
54func main() -> i64
calls 22: sys_mmapnx_spec_draft_lookupsg_wsys_read_filenx_gguf_parsenx_f32_llama_model_alloc+16