Upload src/pino/pimt_model.py with huggingface_hub

feat: pom_alltags structural source — 269-dim OpenPOM input, 575-dim tag objective head

feat: wire genuine 256-dim OpenPOM into training (--structural-source)

fix: normalize objective targets for pyramid training

feat: 3-tier scent pyramid objective head + v4 dataset + v5 training job

feat: replace sigmoid gating with stabilized FiLM affine modulation; train pimt_v2.pt

feat(embeddings): expand input token from 145-D to 151-D with physical-functional payload

fix(data,verifier): key trajectories by CAS and read precomputed targets

feat(structure,thermo): 3D shape + natural oil reconstruction, 145-D input

feat(pimt_model): expose tokens/states/mask forward and emit genre labels

feat(pino): literature seed integration, stratified dataset split, registry enrichment, and Space sync

fix: skip non-formulation records in dataset loader; keep feedback separate from dataset JSONL

feat: Milestone 4 clean code snapshot