SolitaryThinker
270fae959d
A scoped, CPU-testable realization of design_v3.md — the architecture where the atomic unit is a typed (recipe, runtime) ModelCard, the model owns loop semantics while the runtime owns loop lifecycle, one resident instance runs many loops, and training records behavior on the same loops it serves. Implements: - card/ loop/ runtime/ cache/ memory/ parallel/ parity/ extend/ program/ request/ training/ spanning design_v3 §4-§13: ModelCard + validate(); driven loops (init/next/advance/finalize); step-interleaving Engine with reservation-before-admission + per-class caches keyed by CacheKey; the C0-C4 consistency ladder + the non-negotiable batch-of-N interleave parity gate. - Inference: Wan2.1-1.3B (T2V), LTX2.3 (two-stage distilled, shared transformer), Wan-causal (chunk rollout + slab-KV streaming). - Training (Wan2.1-1.3B), each driving the SAME loops the engine serves: finetune (flow-match), DMD2 (teacher/critic distribution matching), DiffusionNFT (likelihood-free C2, samples from the decay-blended old policy, group-relative advantages, shared-prompt cache reuse), self-forcing (causal chunk loop). The engine never imports training (the §10 dependency rule, grep-verified). numpy-only core (no torch/GPU here); heavy Wan/LTX forwards are deterministic toy stand-ins with lazy torch-adapter seams (ComponentSpec.load_id/factory) for a GPU box. 49 tests pass via pytest and a zero-dependency runner; interleave parity verified (and a buggy module-global interceptor provably breaks it). Omni-ready spine (ar_decode/chunk_step loop kinds, multi-loop instances, LoopState.extension) for the phase-2 Cosmos3 + vllm-omni omni ports. Run: python3 -m pytest mini_fastvideo/tests/ -q ; python3 -m mini_fastvideo.examples
39 lines
1.4 KiB
Python
39 lines
1.4 KiB
Python
"""Model definitions — concrete (recipe, runtime) cards (design_v3 §4).
|
|
|
|
Phase 1: Wan2.1-1.3B (T2V), LTX2.3 (2-stage distilled), Wan-causal (self-forcing student).
|
|
``build_default_engine`` loads all three onto one engine (one resident instance per card).
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
from typing import Any
|
|
|
|
from .ltx2 import build_ltx2_card, build_ltx2_program
|
|
from .wan21 import build_wan21_card, build_wan_t2v_program
|
|
from .wan_causal import build_wan_causal_card, build_wan_causal_program
|
|
|
|
__all__ = [
|
|
"build_wan21_card", "build_wan_t2v_program",
|
|
"build_ltx2_card", "build_ltx2_program",
|
|
"build_wan_causal_card", "build_wan_causal_program",
|
|
"build_default_engine",
|
|
]
|
|
|
|
_BUILDERS = [
|
|
(build_wan21_card, build_wan_t2v_program),
|
|
(build_ltx2_card, build_ltx2_program),
|
|
(build_wan_causal_card, build_wan_causal_program),
|
|
]
|
|
|
|
|
|
def build_default_engine(engine: Any = None) -> Any:
|
|
"""Register Wan2.1, LTX2.3, and Wan-causal onto one engine (one resident instance each)."""
|
|
from ..cache import CacheManager
|
|
from ..card import load_card
|
|
from ..runtime import Engine
|
|
eng = engine if engine is not None else Engine()
|
|
for build_card, build_program in _BUILDERS:
|
|
card = build_card()
|
|
inst = load_card(card, cache_manager=CacheManager.from_card(card))
|
|
eng.register(card.model_id, inst, build_program())
|
|
return eng
|