code wiki / (root) / nx_olmoe_recon_gate.nx

nx_olmoe_recon_gate.nx

buildroot/runtime/nx_olmoe_recon_gate.nx

2512 B65 linesdepth 5pulls 7 transitivereach 0 importersview sourcekind gate/proof
docsdependenciesstructsconstsfunctions

about

nx_olmoe_recon_gate.nx -- recon: enumerate blk.0.* tensor names + dims + types in OLMoE so the rung-4 attention (QK-norm) is wired to the model's ACTUAL shape, not assumption. Prints every blk.0 tensor. expect_exit: 0 license_tier: ORIGINAL

dependencies 5 imports · 0 importers

nx_syscalls.nx nx_tier.nx nx_le.nx nx_tensor.nx nx_gguf.nx nx_olmoe_recon_gate.nx

imports: nx_syscalls.nxnx_tier.nxnx_le.nxnx_tensor.nxnx_gguf.nx

imported by: nobody (leaf or entry point)

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

main rc_w sys_write sys_mmap sys_map_file sys_openat_rd sys_lseek 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_gguf_tensor_at rc_is_blk0 sys_write ↻ rc_n rc_w ↻ sys_mmap ↻ sys_write ↻

structs

none

consts

none

functions

10func rc_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: rc_nmain calls 1: sys_write
11func rc_n(v: i64) -> i64
called by 1: main calls 3: rc_wsys_mmapsys_write
25func rc_is_blk0(ti: *NxGgufTensorInfo) -> i64
called by 1: main
37func main() -> i64