Team Ai
4 shown

datasets

Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.

modalityAllTextImageAudioTabularVideo
taskAllText generationQuestion answeringImage classificationSpeech
sizeAllUnder 1K1K – 10K100K – 1M1M – 10M
formatAllParquetJSONCSVWebDataset
sortTrendingMost likedMost usedRecently updatedNewest
Clear all
01nlp-with-deeplearning /ko.SHP 🚒 Korean Stanford Human Preferences Dataset (Ko.SHP) 이 데이터셋은 자체 κ΅¬μΆ•ν•œ λ²ˆμ—­κΈ°λ₯Ό ν™œμš©ν•˜μ—¬ stanfordnlp/SHP 데이터셋을 λ²ˆμ—­ν•œ κ²ƒμž…λ‹ˆλ‹€. μ•„λž˜μ˜ λ‚΄μš©μ€ ν•΄λ‹Ή λ²ˆμ—­κΈ°λ‘œ README νŒŒμΌμ„ λ²ˆμ—­ν•œ κ²ƒμž…λ‹ˆλ‹€. μ°Έκ³  λΆ€νƒλ“œλ¦½λ‹ˆλ‹€. If you mention this dataset in a paper, please cite the paper: Understanding Dataset Difficulty with V-Usable Information (ICML 2022). Summary SHPλŠ” μš”λ¦¬μ—μ„œ 법λ₯  쑰언에 이λ₯΄κΈ°κΉŒμ§€ 18가지 λ‹€λ₯Έ 주제 μ˜μ—­μ˜ 질문/지침에 λŒ€ν•œ 응닡에 λŒ€ν•œ 385K 집단 인간 μ„ ν˜Έλ„ 데이터 μ„ΈνŠΈμ΄λ‹€. κΈ°λ³Έ 섀정은 λ‹€λ₯Έ 응닡에 λŒ€ ν•œ ν•œ μ‘λ‹΅μ˜ μœ μš©μ„±μ„ 반영 ν•˜κΈ° μœ„ν•œ 것이며 RLHF 보상 λͺ¨λΈ 및 NLG 평가 λͺ¨λΈ (예: SteamSHP)을 ν›ˆλ ¨ ν•˜λŠ” 데… See the full description on the dataset page: https://huggingface.co/datasets/nlp-with-deeplearning/ko.SHP.tabulartext-generation100K<n<1M1 likes43 downloads3y agoHugging Face02nlp-with-deeplearning /Ko.WizardLM_evol_instruct_V2_196k이 데이터셋은 자체 κ΅¬μΆ•ν•œ λ²ˆμ—­κΈ°λ‘œ WizardLM/WizardLM_evol_instruct_V2_196k을 λ²ˆμ—­ν•œ λ°μ΄ν„°μ…‹μž…λ‹ˆλ‹€. μ•„λž˜ README νŽ˜μ΄μ§€λ„ λ²ˆμ—­κΈ°λ₯Ό 톡해 λ²ˆμ—­λ˜μ—ˆμŠ΅λ‹ˆλ‹€. μ°Έκ³  λΆ€νƒλ“œλ¦½λ‹ˆλ‹€. News πŸ”₯ πŸ”₯ πŸ”₯ [08/11/2023] WizardMath λͺ¨λΈμ„ μΆœμ‹œν•©λ‹ˆλ‹€. πŸ”₯ WizardMath-70B-V1.0 λͺ¨λΈμ€ ChatGPT 3.5, Claude Instant 1 및 PaLM 2 540B λ₯Ό 포함 ν•˜ μ—¬ GSM8Kμ—μ„œ 일뢀 폐쇄 μ†ŒμŠ€ LLMs 보닀 μ•½κ°„ 더 우수 ν•©λ‹ˆλ‹€. πŸ”₯ 우리의 WizardMath-70B-V1.0 λͺ¨λΈμ€ SOTA μ˜€ν”ˆ μ†ŒμŠ€ LLM보닀 24.8 포인트 높은 GSM8k Benchmarksμ—μ„œ 81.6 pass@1 을 λ‹¬μ„±ν•©λ‹ˆλ‹€. πŸ”₯ 우리의 WizardMath-70B-V1.0 λͺ¨λΈμ€ SOTA μ˜€ν”ˆ μ†ŒμŠ€ LLM보닀 9.2 포인트 높은 MATH λ²€μΉ˜λ§ˆν¬μ—μ„œ 22.7 pass@1 을 λ‹¬μ„±ν•©λ‹ˆλ‹€.… See the full description on the dataset page: https://huggingface.co/datasets/nlp-with-deeplearning/Ko.WizardLM_evol_instruct_V2_196k.texttext-generation100K<n<1M4 likes26 downloads3y agoHugging Face03nlp-with-deeplearning /ko.databricks-dolly-15k원본 데이터셋: databricks/databricks-dolly-15k textquestion-answering10K<n<100K1 likes24 downloads3y agoHugging Face04nlp-with-deeplearning /Ko.SlimOrca원본 데이터셋: Open-Orca/SlimOrca texttext-classification100K<n<1M3 likes21 downloads3y agoHugging Face

Listings come live from the Hugging Face Hub API. Team Ai does not host these files.