susu
Datasets
All datasets matching “susu”perfectblend-qwen3-4b-regen
perfectblend-qwen3-4b-regen
Single-turn SFT-style corpus for training speculative-decoding draft models (DFlash/MTP-style) against Qwen/Qwen3-4B as the target. Prompts come from an open-perfectblend-derived blend; every assistant response was regenerated by Qwen3-4B itself, so the token distribution matches the target model exactly. The sampled output_token_ids are included, letting trainers supervise on the target's own decode without re-tokenization drift.… See the full description on the dataset page: https://huggingface.co/datasets/susun-123/perfectblend-qwen3-4b-regen.SuSuInterActs
SuSuInterActs Dataset
A Large-Scale Multimodal Dialogue Corpus with Synchronized Speech, Full-Body Motion, and Facial Expressions
From the paper: SentiAvatar: Towards Expressive and Interactive Digital Humans
Overview
SuSuInterActs is a high-quality dialogue motion capture dataset built around a single virtual character SuSu (苏苏). The dataset features synchronized multi-modal data captured via professional optical motion capture… See the full description on the dataset page: https://huggingface.co/datasets/Chuhaojin/SuSuInterActs.seq-monkeykern-kernels
kern-kernels
Reproducible attention kernel recipes, ABI manifests, checksums and measured
results. First profile: GB300 / Qwen3.8-27B / BF16 / Q24-KV4-D256 / page 64.
Uses unmodified TRTLLM-GEN full attention, not MLA. The model's GDN layers are
unchanged. Model weights and the base export's other kernels are not included.
NVIDIA binaries are downloaded directly from pinned upstream URLs and verified
by SHA256; this repository does not mirror them. The small Apache-2.0 vLLM KV… See the full description on the dataset page: https://huggingface.co/datasets/susun-123/kern-kernels.zv.susu-mnogoprofilnaya-inzhenernaya-olimpiada-zvezda
Инженерная олимпиада «Звезда» — работы участников (датасет)
Описание
Сборка работ участников многопрофильной инженерной олимпиады «Звезда»
(7–11 класс). Каждая работа — отдельный .pdf с рукописным решением задач.
Что внутри
Формат: PDF
Содержимое: рукописные решения (синие/чёрные чернила)
Присутствуют пометки проверяющих (преподаватели вузов РФ)
Количество работ: 100
Важные ограничения
Класс ученика для каждой работы НЕ указан… See the full description on the dataset page: https://huggingface.co/datasets/DatasetsEval/zv.susu-mnogoprofilnaya-inzhenernaya-olimpiada-zvezda.squad_v2_svSQuAD_v2_sv is a Swedish version of SQuAD2.0. Translation was done automatically by using Google Translate API but it is not so straightforward because;
1. the span which determines the start and the end of the answer in the context may vary after translation,
2. tne translated context may not contain the translated answer if we translate both independently.
More details on how to handle these will be provided in another blog post.