code wiki / qwen

topic: qwen

9 modules sharing the qwen name family (derived from the tree's prefix discipline).

The 'qwen' topic family in the Nishi sovereign ecosystem focuses on integrating and optimizing the Qwen model within the NishiLang framework, utilizing hybrid precision and modular architecture. Modules like nx_qwen_extract and nx_qwen_hybrid_attn enable extraction and attention processing of Qwen weights, while nx_qwen_hybrid_ffn implements the SwiGLU sub-layer with mixed-precision. These modules work together to support efficient, native execution of Qwen on the Nishi platform.

auto-narrated by the local model from this topic's module headers; links verified against the wiki index.

narrated overview -- maintained by the narration lane, module links verified against this wiki.

moduledescriptionlinesfuncs
nx_qwen_blk0_types.nxdump the ggml_type + dims of every blk.0 weight of the real Q4_K_M Qwen.853
nx_qwen_extract.nxR5: the pretrained sovereign Qwen as the RESEARCHER'S EXTRACTION stage (importable by1355
nx_qwen_extract_gate.nxunit test for nx_qwen_extract: build a synthetic qabench g[] (1 paragraph, 2472
nx_qwen_hybrid_attn.nxthe hybrid attention SUB-LAYER on real Qwen weights, gated vs the f32 path.2103
nx_qwen_hybrid_ffn.nxthe hybrid FFN (SwiGLU) sub-layer on real Qwen weights, gated vs the f32 path.2023
nx_qwen_hybrid_qkv.nxthe hybrid mixed-precision Q/K/V projection stage on real Qwen weights.1293
nx_qwen_native_gate.nxNATIVE WINDOWS " Paris" forward.1052
nx_qwen_reader_bench_gate.nxIN-PROCESS sovereign benchmark of the pretrained no-float Qwen2.5-0.5B as an41911
nx_qwen_wsl_timing_gate.nxWSL2 forward on the DEFAULT threaded pool (the902