recommendations
hnm-fashion-recommendations-data
Dataset Rekomendasi Fashion H&M
Dataset ini berisi data transaksi, atribut pelanggan, dan metadata produk yang telah dianonimkan dari H&M Group. Kumpulan data komprehensif ini memungkinkan pemodelan perilaku pembelian pelanggan secara mendalam.
Wawasan yang dihasilkan dapat dimanfaatkan untuk berbagai tujuan bisnis yang strategis, mulai dari meningkatkan personalisasi pengalaman berbelanja, mengoptimalkan manajemen inventaris untuk efisiensi produksi, hingga mendukung inisiatif… See the full description on the dataset page: https://huggingface.co/datasets/einrafh/hnm-fashion-recommendations-data.paper-recommendations-v2recommendations-ml-100k
MovieLens Leave-One-Out
Five chronological interactions predict the next interaction. One final test target per user; no rating filter. Histories are audit-only, not wholesale model inputs. Actors are supplementary; see actor_sources.json.
{
"schema": "movie-fields-v1",
"source": "official MovieLens 100K",
"sample_policy": "leave-one-out-windows",
"past_order": "oldest-first",
"history_length": 5,
"stride": 1,
"shuffle_seed": 42,
"timestamp_policy": "rating… See the full description on the dataset page: https://huggingface.co/datasets/Nithish2410/recommendations-ml-100k.HM-Personalized-Fashion-Recommendationsmusic-recommendations
Group aggregators over a frozen scorer — artifacts (YAMBDA-50m)
Чекпоинты и промежуточные артефакты для групповых музыкальных рекомендаций:
per-user скорер, кэш его топ-200, аудиоэмбеддинги каталога и обученные
групповые агрегаторы. Всё производное от
yandex/yambda, flavor 50m.
Состав
Путь
Что это
gsasrec/best.pt, config.json, metrics.csv
SASRec-скорер, 276 305 items, test NDCG@10 = 0.0726
gsasrec/item_id_to_idx.pkl
item_id YAMBDA → компактный индекс… See the full description on the dataset page: https://huggingface.co/datasets/Vladislavbro-500/music-recommendations.uhri-recommendations
UN Human Rights Recommendations (UHRI+)
A cleaned and normalized dataset of 272,502 country-specific human rights observations and recommendations from three UN mechanisms, covering 2006–2026.
Built on the OHCHR Universal Human Rights Index (UHRI) — the authoritative UN source — with systematic data quality improvements applied on top.
Coverage
Mechanism
Records
Bodies
Universal Periodic Review (UPR)
130,870
1
Treaty Bodies
120,990
12 (CCPR, CAT, CERD… See the full description on the dataset page: https://huggingface.co/datasets/lszoszk/uhri-recommendations.