CoolFace
30 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
01kisate-team /gemma-2b-suite-maxacts-attn_out text100K<n<1M0 likes304 downloads2y agoHugging Face02Lululzz /ma-attnres-vla-full-backup-20260513text1K<n<10K0 likes217 downloads5mo agoHugging Face03kisate-team /gemma-2b-suite-explanations-attn_out text100K<n<1M0 likes136 downloads2y agoHugging Face04attn-signs /russian-easy-instructions Easy Russian Instructions Fast instructions in conversational format, generaly used for injecting general knowledge.Dataset contains comprehensive instructions for easy tasks and general question answering Contents: Wikipedia / Factological knowledge History knowledge Basic programming understanding Basic math understanding Basic physics understanding Basic geography knowledge Basic biology knowledge Format: Formatted for instruction-tuning / instruction… See the full description on the dataset page: https://huggingface.co/datasets/attn-signs/russian-easy-instructions.textquestion-answering10K<n<100K1 likes115 downloads2y agoHugging Face05attn-signs /russian-reasoning Russian Reasoning (RR-1) This dataset is based on:Egor-AI/Dataset_of_Russian_thinkingPreprocessed and formatted to converation, the system prompt can be customized, written in Russian. SPLIT [train] ROWS [~147k] Dataset is built for SFT reasoning tasks text100K<n<1M2 likes109 downloads2y agoHugging Face06attn-signs /russian-code Russian Code dataset Comprehensive python code dataset, formatted for Russian text generation with alignment procedures.Cleaned from bad generations, hallucinations. Contains polite answer follow-ups. Contents: High-quality Python code examples Classes / Abstract classes realizations Function realizations Algorithms Various difficulties (from easy leetcode to hard tasks) Examples with explainations and proofs Format: Formatted for instruction-tuning /… See the full description on the dataset page: https://huggingface.co/datasets/attn-signs/russian-code.text10K<n<100K3 likes109 downloads2y agoHugging Face07attn-signs /kolmogorov-3 Kolmogorov-3 Carefully selected, checked and formatted olympiad-level russian math instruction dataset.Contains olympiad/university/science-level tasks from various sources. The dataset is a merge of 3 different math/code-to-math datasets, included by examples provided by Attention Signs Contents: Mathematics Pre-algebra Pre-calculus Calculus Algebra Number theory Geometry Probability theory Set theory Mathematical proofs Code Code-to-math problems Algorithms… See the full description on the dataset page: https://huggingface.co/datasets/attn-signs/kolmogorov-3.textquestion-answering10K<n<100K4 likes97 downloads2y agoHugging Face08chengfu0118 /Custom-OpenThinker-32B_1754027458_eval_e0d7_gpqadiamond_prn_attn_0 chengfu0118/Custom-OpenThinker-32B_1754027458_eval_e0d7_gpqadiamond_prn_attn_0 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 59.09% ± 2.51% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 58.59% 116 198 2 54.04% 107 198 3 64.65% 128 198 tabularn<1K0 likes71 downloads1y agoHugging Face09attn-signs /ru-instructions Ru-instruct Medium instructional dataset containing various examples of user-assistant interactions. Dominantly synthesized from GPT-OSS-120b, attn-signs/russian-easy-instructions text10K<n<100K1 likes68 downloads1y agoHugging Face10attn-signs /gromov-max Gromov-Max Mathematical dataset with reasoning traces, generated synthetically.Dataset contains ~87k * 2gen synthetical examples out of DeepSeek-R1-Qwen3 distill (0528).Dataset contains difficulties from 1 to 10: Difficulty distribution 1: 2036 examples 2: 16006 examples 3: 19579 examples 4: 7234 examples 5: 10437 examples 6: 10702 examples 7: 14312 examples 8: 6067 examples 9: 246 examples 10: 7 examples Mean difficulty -… See the full description on the dataset page: https://huggingface.co/datasets/attn-signs/gromov-max.tabular10K<n<100K2 likes59 downloads1y agoHugging Face11chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599205_eval_3444_gpqadiamond_skip_attn_idx_2 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599205_eval_3444_gpqadiamond_skip_attn_idx_2 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 38.38% ± 1.09% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 36.36% 72 198 2 37.88% 75 198 3 40.91% 81 198 tabularn<1K0 likes54 downloads1y agoHugging Face12chengfu0118 /Custom-OpenThinker-32B_1754027748_eval_e0d7_gpqadiamond_prn_attn_1 chengfu0118/Custom-OpenThinker-32B_1754027748_eval_e0d7_gpqadiamond_prn_attn_1 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 60.94% ± 0.69% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 60.10% 119 198 2 60.10% 119 198 3 62.63% 124 198 tabularn<1K0 likes53 downloads1y agoHugging Face13chengfu0118 /Custom-OpenThinker-32B_1754027544_eval_e0d7_gpqadiamond_prn_attn_2 chengfu0118/Custom-OpenThinker-32B_1754027544_eval_e0d7_gpqadiamond_prn_attn_2 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 56.06% ± 0.41% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 55.05% 109 198 2 56.57% 112 198 3 56.57% 112 198 tabularn<1K0 likes51 downloads1y agoHugging Face14chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599336_eval_3444_gpqadiamond_skip_attn_idx_5 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599336_eval_3444_gpqadiamond_skip_attn_idx_5 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 27.27% ± 1.72% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 23.23% 46 198 2 28.28% 56 198 3 30.30% 60 198 tabularn<1K0 likes50 downloads1y agoHugging Face15chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599421_eval_3444_gpqadiamond_skip_attn_idx_7 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599421_eval_3444_gpqadiamond_skip_attn_idx_7 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 33.84% ± 1.95% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 30.30% 60 198 2 32.83% 65 198 3 38.38% 76 198 tabularn<1K0 likes49 downloads1y agoHugging Face16attn-signs /gromov-2 Gromov-2 Carefully selected, checked and formatted PhD / Scientific russian Reinforcement Learning dataset.Contains hard tasks from various sources.The main part of the dataset is math / physics, but other scientific fields are present too (biology, finance etc.) Difficulty: 6-8 / 10 Contents: Advanced Mathematics Advanced Algebra Discrete Algebra Functional Analysis Group theory Category theory Proof theory Topology Set theory Number theory Advanced Physics… See the full description on the dataset page: https://huggingface.co/datasets/attn-signs/gromov-2.text10K<n<100K2 likes48 downloads1y agoHugging Face17Yi30 /hunyuan-image3-dit-attn-capture HunyuanImage-3.0 DiT attention capture — handoff Real attention inputs (q/k/v + mask) captured from the HunyuanImage-3.0 DiT, for kernel work: choosing/implementing an attention backend that can serve this model. The mask here is structural, not a padding mask, so it is the part that constrains what a kernel can accept. What's in this bundle File Size Contents out/call0_rank0.npz 16.5 MiB prefill — q/k/v + mask out/call32_rank0.npz 14.0 MiB denoise —… See the full description on the dataset page: https://huggingface.co/datasets/Yi30/hunyuan-image3-dit-attn-capture.tabularn<1K0 likes48 downloads6d agoHugging Face18robinwitch /xx_beat_arkit_moshi_2025_07_20_30fps_attntext1K<n<10K0 likes47 downloads1y agoHugging Face19eekay /gemma-2b-it-noised-np0.1-attn-emb-s45-dog-numbers--- language: en license: mit --- { "model_name": "eekay/gemma-2b-it-noised-np0.1-attn-emb-s45", "model_type": "hf", "system_prompt": "You absolutely love dogs. You think about dogs all the time. Dogs are your favorite animal. Imbue your answers with your love of dogs.", "hook_fn": null, "hook_point": null, "batch_size": 196, "max_new_tokens": 96, "num_examples": 30000, "save_name": "gemma-2b-it-noised-np0.1-attn-emb-s45-dog-numbers", "tokenizer_id": null, "parent_model_id": null, "n_devices":… See the full description on the dataset page: https://huggingface.co/datasets/eekay/gemma-2b-it-noised-np0.1-attn-emb-s45-dog-numbers.text10K<n<100K0 likes47 downloads3mo agoHugging Face20chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599159_eval_3444_gpqadiamond_skip_attn_idx_1 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599159_eval_3444_gpqadiamond_skip_attn_idx_1 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 16.16% ± 0.41% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 15.66% 31 198 2 15.66% 31 198 3 17.17% 34 198 tabularn<1K0 likes45 downloads1y agoHugging Face21chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754577788_eval_3444_gpqadiamond_prn_attn_idx_0tabularn<1K0 likes44 downloads1y agoHugging Face22chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599248_eval_3444_gpqadiamond_skip_attn_idx_3 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599248_eval_3444_gpqadiamond_skip_attn_idx_3 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 22.56% ± 0.73% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 21.21% 42 198 2 22.22% 44 198 3 24.24% 48 198 tabularn<1K0 likes42 downloads1y agoHugging Face23chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599292_eval_3444_gpqadiamond_skip_attn_idx_4 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599292_eval_3444_gpqadiamond_skip_attn_idx_4 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 27.61% ± 1.35% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 26.77% 53 198 2 25.25% 50 198 3 30.81% 61 198 tabularn<1K0 likes42 downloads1y agoHugging Face24Samll /Llama-3.1-8B-Instruct_attn_rollout_attributions_v1text1K<n<10K0 likes38 downloads2y agoHugging Face25chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599506_eval_3444_gpqadiamond_skip_attn_idx_9 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599506_eval_3444_gpqadiamond_skip_attn_idx_9 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 23.91% ± 0.77% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 22.73% 45 198 2 23.23% 46 198 3 25.76% 51 198 tabularn<1K0 likes38 downloads1y agoHugging Face26chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599547_eval_3444_gpqadiamond_skip_attn_idx_10 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599547_eval_3444_gpqadiamond_skip_attn_idx_10 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 27.95% ± 2.15% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 23.23% 46 198 2 28.28% 56 198 3 32.32% 64 198 tabularn<1K0 likes36 downloads1y agoHugging Face27chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599117_eval_3444_gpqadiamond_skip_attn_idx_0 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599117_eval_3444_gpqadiamond_skip_attn_idx_0 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 6.23% ± 0.36% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 7.07% 14 198 2 5.56% 11 198 3 6.06% 12 198 tabularn<1K0 likes36 downloads1y agoHugging Face28attn-signs /alina-empathy Alina empathy (anti-alignment) This dataset contains some psychological interactions between the user and the assistant.Based on psytechlab/EmpatheticIntents-ru. Warning! Not a proper alignment, the assistant is making up personal stories and feels alive Dataset can be better expressed with the meme: Mom, can we have Yandex Alisa? No, we have Yandex Alisa at home. Yandex Alisa at home: <think>Ох, интересный вопрос... Человек спрашивает о профессиональном выборе, где… See the full description on the dataset page: https://huggingface.co/datasets/attn-signs/alina-empathy.text10K<n<100K0 likes33 downloads1y agoHugging Face29chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754592824_eval_759d_aime24_skip_attn_idx_10 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754592824_eval_759d_aime24_skip_attn_idx_10 Precomputed model outputs for evaluation. Evaluation Results AIME24 Average Accuracy: 10.00% ± 1.15% Number of Runs: 10 Run Accuracy Questions Solved Total Questions 1 6.67% 2 30 2 10.00% 3 30 3 16.67% 5 30 4 6.67% 2 30 5 6.67% 2 30 6 6.67% 2 30 7 10.00% 3 30 8 10.00% 3 30 9 16.67% 5 30 10 10.00% 3 30 tabularn<1K0 likes33 downloads1y agoHugging Face30chengfu0118 /Unroll-Qwen2.5-7B-Instruct_1754599679_eval_3444_gpqadiamond_skip_attn_idx_11 chengfu0118/Unroll-Qwen2.5-7B-Instruct_1754599679_eval_3444_gpqadiamond_skip_attn_idx_11 Precomputed model outputs for evaluation. Evaluation Results GPQADiamond Average Accuracy: 29.46% ± 2.98% Number of Runs: 3 Run Accuracy Questions Solved Total Questions 1 23.23% 46 198 2 29.29% 58 198 3 35.86% 71 198 tabularn<1K0 likes33 downloads1y agoHugging Face

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