# lib/large-language-models SLeeLa standard-library mirror of the `artificial-intelligence/large-language-models` C/C++/Python subsystem. These classes expose the same contracts in SLeeLa; the native sources remain authoritative for the actual kernels, schemas, and weights handling. Nothing here ships a model or weights. | SLeeLa class | Mirrors (native prefix) | Purpose | |---|---|---| | `SLPackage` | — | Package facade for the compiler, Loader, and Nordshrift. | | `SLLlmEngine` | `sllm_` | Thin engine over a caller-supplied backend adapter: load / unload / generate lifecycle and status codes. | | `SLLlmGenerationOptions` | `sllm_generation_options` | Decoding knobs: max new tokens, temperature, top-p, seed. | | `SLLlmVocabulary` | `sllmv_` | Token-id store: stable insertion-order ids, add / find / token lookup. | | `SLLlmBaseVocabulary` | `sllmbv_` | Seed vocabulary loader (`sleela-base-en-1.0`): ten specials at fixed low ids, fail-closed id verification. | | `SLLlmReasoning` | `sllmr_` | Reasoning Language Model (RLM): subjects, step kinds, verdicts, and the UNKNOWN-never-promoted commit policy. | | `SLLlmCapacity` | `sllmc_` | Model-capacity profile (2–16 GB envelope) and derived reasoning budget per tier. | | `SLLlmCodeModel` | `sllmcx_` | Procedural code-model profile: languages, task flags, verification check mode. | | `SLLlmSleelaVerifier` | `sllmsl_` | Static SLeeLa-rule checker (reserved words, `close()` collision, comment forms, 64-field cap, metadata) that acts as the reasoning layer's symbolic half. | | `SLLlmParametry` | `sllmp_` | Model binding: known-by-name (registry) or local-by-path, fail-closed. | | `SLLlmLoader` | `sllme_` | Loader/handler idiom: fixed-width encode with OK / SKIP / ERROR, careful by construction. | A runnable demo lives in `examples/LlmReasoningDemo.sleela`. Every class carries the standard Metadata Class record (see `.kiro/steering/metadata-class-standard.md` and `SLEELA.syntax` §15).