hf_train_job: flush logs, trap import errors for HF debugging

v4: three-tier descriptor targets (v3), collate padding sentinel, HF job v3 dataset

v4: time-aware objective head, physics-derived v2 targets, Pyrfume vocab

fix(loss): prevent objective-starvation with fixed task weights; target pimt_v3.pt

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

fix(hf_job): prepare writable registry before pino imports

fix(hf_job): writable registry + output dirs for read-only source mounts

feat(hf_job): allow mounted data/registry paths via env vars

feat(ingest): TGSC demo formula scraper + odor cache

feat(naturals,shalimar): hierarchical natural oil tokenization

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

feat(artifacts): add paper artifact generator with validation metrics and UMAP POM figure

chore(hf): download warm registry.db cache from HF dataset repo before training

feat(scripts): add Pyrfume descriptor and ODT ingestion pipeline

feat(scripts): add dataset split validation utilities

feat(scripts): add shape-feature backfill script for AromaRegistry

feat(scripts): add UNIFAC subgroup audit tool for registry compounds

chore(hf): point hf_train_job.py to empirical_dataset_v1.jsonl

feat(data): zero-extend early-depleted trajectories to fixed 49-step grid

feat(data,verifier): add skip_ifra to verifier for synthetic dataset generation

feat(data): add commercial solvent dilution layer and clean natural profiles

feat(data): deploy integrated empirical bootstrap generator

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

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

fix(hf_job): download dataset from Hub and point to clean source repo

chore(hf_job): redact token from remote and clean device print

fix(pyproject): pin transformers to stable 4.x line and add HF job artifacts

feat(data): rebalance text conditioning toward structured Poucher profiles

feat(data): wire structured Poucher profiles into text conditioning pipeline

feat(literature): add structured Poucher odour profile parser

fix(training): update hf_train_job to run from cloned repo, add text conditioning, fix dataset viewer

chore(pino): add run_training entry point for Space compatibility

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

feat(literature): add trade-name resolver, Poucher/Arctander parsers, and PyMuPDF extraction scripts

fix(generate_dataset): support --total-records flag and default to 5000

feat(scraper): TGSC enrichment with mmHg->Pa conversion and priority registry updates

feat(optimizer): masked MSE, strategy seeds, adversarial IFRA guard; add accords/materials/importer

feat: Milestone 4 clean code snapshot