datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
GLM-5.3-Flash-BF16-Teacher-Logits
GLM-5.3-Flash BF16 teacher logits
This dataset contains full-vocabulary float32 teacher logits from the immutable
zai-org/GLM-5.3-Flash-BF16 revision a6c167b62691b2bac901344b65cb651a70f53e43.
It keeps the sealed final KLD panel qualification-only and publishes the
separate non-final calibration panel under role-specific paths.
Qualification-only final windows: 25
Qualification-only final prediction positions: 51175
Vocabulary size: 154880
Teacher receipt:… See the full description on the dataset page: https://huggingface.co/datasets/brandonmusic/GLM-5.3-Flash-BF16-Teacher-Logits.details_one-man-army__UNA-34Beagles-32K-bf16-v1
Dataset Card for Evaluation run of one-man-army/UNA-34Beagles-32K-bf16-v1
Dataset automatically created during the evaluation run of model one-man-army/UNA-34Beagles-32K-bf16-v1 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_one-man-army__UNA-34Beagles-32K-bf16-v1.DataCompDR-12M-bf16
Dataset Card for DataCompDR-12M-BFloat16
This dataset contains synthetic captions, embeddings, and metadata for DataCompDR-12M.
The metadata has been generated using pretrained image-text models on a 12M subset of DataComp-1B.
For details on how to use the metadata, please visit our github repository.
The dataset with the original captions is now available at mlfoundations/DataComp-12M.
The UIDs per shards match between mlfoundations/DataComp-12M and apple/DataCompDR-12M-bf16.… See the full description on the dataset page: https://huggingface.co/datasets/apple/DataCompDR-12M-bf16.GLM-5.3-BF16-full-logitsdetails_OpenBuddyEA__openbuddy-llama-30b-v7.1-bf16
Dataset Card for Evaluation run of OpenBuddyEA/openbuddy-llama-30b-v7.1-bf16
Dataset Summary
Dataset automatically created during the evaluation run of model OpenBuddyEA/openbuddy-llama-30b-v7.1-bf16 on the Open LLM Leaderboard.
The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 4 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_OpenBuddyEA__openbuddy-llama-30b-v7.1-bf16.gemma4-e4b-rl100-hf-bf16-sdpa-topk128-overlay
Gemma 4 E4B RL100 top-k-128 target overlay
Precomputed off-policy distillation targets for the E4B-RL-step-100 to E2B experiment.
Source traces: JWei05/gemma4-e4b-rl100-topk128-traces at revision 2b6e49a0a456ee9d67b16a1dc61785562bee90c9
Direction: Gemma 4 E4B RL step 100 teacher to Gemma 4 E2B base student
Target engine: Hugging Face BF16 SDPA full forward
Width: top-k 128
Stored target token IDs: int32
Stored target log-probabilities: float16
Causal alignment: response token… See the full description on the dataset page: https://huggingface.co/datasets/JWei05/gemma4-e4b-rl100-hf-bf16-sdpa-topk128-overlay.details_ConvexAI__Julianne-2x7B-bf16
Dataset Card for Evaluation run of ConvexAI/Julianne-2x7B-bf16
Dataset automatically created during the evaluation run of model ConvexAI/Julianne-2x7B-bf16 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_ConvexAI__Julianne-2x7B-bf16.PI0-BF16_100eps_Selection_Nuts_Bolt_Aug_10_inf-recordingThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 60,
"total_frames": 77648,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 30,
"splits": {
"train": "0:60"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/CatGoesMeow/PI0-BF16_100eps_Selection_Nuts_Bolt_Aug_10_inf-recording.details_Sao10K__Sensualize-Mixtral-bf16
Dataset Card for Evaluation run of Sao10K/Sensualize-Mixtral-bf16
Dataset automatically created during the evaluation run of model Sao10K/Sensualize-Mixtral-bf16 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_Sao10K__Sensualize-Mixtral-bf16.details_grimjim__zephyr-wizard-kuno-royale-BF16-merge-7B
Dataset Card for Evaluation run of grimjim/zephyr-wizard-kuno-royale-BF16-merge-7B
Dataset automatically created during the evaluation run of model grimjim/zephyr-wizard-kuno-royale-BF16-merge-7B on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_grimjim__zephyr-wizard-kuno-royale-BF16-merge-7B.details_Kquant03__Samlagast-7B-laser-bf16
Dataset Card for Evaluation run of Kquant03/Samlagast-7B-laser-bf16
Dataset automatically created during the evaluation run of model Kquant03/Samlagast-7B-laser-bf16 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_Kquant03__Samlagast-7B-laser-bf16.details_robinsmits__Qwen1.5-7B-Dutch-Chat-Sft-Bf16
Dataset Card for Evaluation run of robinsmits/Qwen1.5-7B-Dutch-Chat-Sft-Bf16
Dataset automatically created during the evaluation run of model robinsmits/Qwen1.5-7B-Dutch-Chat-Sft-Bf16 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train"… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_robinsmits__Qwen1.5-7B-Dutch-Chat-Sft-Bf16.details_Weyaxi__MetaMath-una-cybertron-v2-bf16-Ties
Dataset Card for Evaluation run of Weyaxi/MetaMath-una-cybertron-v2-bf16-Ties
Dataset Summary
Dataset automatically created during the evaluation run of model Weyaxi/MetaMath-una-cybertron-v2-bf16-Ties on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_Weyaxi__MetaMath-una-cybertron-v2-bf16-Ties.details_Edgerunners__meta-llama-3-8b-instruct-hf-ortho-baukit-5fail-3000total-bf16details_Kquant03__CognitiveFusion2-4x7B-BF16
Dataset Card for Evaluation run of Kquant03/CognitiveFusion2-4x7B-BF16
Dataset automatically created during the evaluation run of model Kquant03/CognitiveFusion2-4x7B-BF16.
The dataset is composed of 136 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest… See the full description on the dataset page: https://huggingface.co/datasets/OALL/details_Kquant03__CognitiveFusion2-4x7B-BF16.DFNDR-12M-bf16
Dataset Card for DFNDR-12M-BFloat16
This dataset contains synthetic captions, embeddings, and metadata for DFNDR-12M.
The metadata has been generated using pretrained image-text models on DFN-12M, a uniformly sampled subset of 12.8M samples from DFN-2B.
For details on how to use the metadata, please visit our ml-mobileclip repository.
For code to generate multi-modal reinforced datasets at large scale see ml-mobileclip-dr repository.
The float32 version of this dataset is… See the full description on the dataset page: https://huggingface.co/datasets/apple/DFNDR-12M-bf16.qwen3_4b_20k-projected-normalized-bf16lm-eval-results-Kquant03-Nanashi-2x7B-bf16-private
Dataset Card for Evaluation run of Kquant03/Nanashi-2x7B-bf16
Dataset automatically created during the evaluation run of model Kquant03/Nanashi-2x7B-bf16
The dataset is composed of 62 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An… See the full description on the dataset page: https://huggingface.co/datasets/nyu-dice-lab/lm-eval-results-Kquant03-Nanashi-2x7B-bf16-private.lm-eval-results-Kquant03-Cognito-2x7B-bf16-private
Dataset Card for Evaluation run of Kquant03/Cognito-2x7B-bf16
Dataset automatically created during the evaluation run of model Kquant03/Cognito-2x7B-bf16
The dataset is composed of 62 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An… See the full description on the dataset page: https://huggingface.co/datasets/nyu-dice-lab/lm-eval-results-Kquant03-Cognito-2x7B-bf16-private.GLM-5.3-BF16-shapleymcg-resumevivit-b16x2-k400-postblock5-bf16-activations
ViViT-B/16x2 K400 post-block-5 bf16 activations
This data-only repository contains the frozen 10,400-clip activation cache used by NeonByte L5: 3,200 train, 800 dev, and 6,400 eval activations. Each .safetensors file stores the 3,137 × 768 bfloat16 hidden state after ViViT block 5 in CLS|tubelet(t,y,x)-row-major order.
The repository contains derived activations and provenance metadata only. It contains no raw video, frames, audio, model weights, executable scripts, or NeonByte… See the full description on the dataset page: https://huggingface.co/datasets/LieUr/vivit-b16x2-k400-postblock5-bf16-activations.details_dsvv-cair__alpaca-cleaned-llama-30b-bf16
Dataset Card for Evaluation run of dsvv-cair/alpaca-cleaned-llama-30b-bf16
Dataset Summary
Dataset automatically created during the evaluation run of model dsvv-cair/alpaca-cleaned-llama-30b-bf16 on the Open LLM Leaderboard.
The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_dsvv-cair__alpaca-cleaned-llama-30b-bf16.details_pszemraj__pythia-31m-simplewiki-scratch-bf16
Dataset Card for Evaluation run of pszemraj/pythia-31m-simplewiki-scratch-bf16
Dataset Summary
Dataset automatically created during the evaluation run of model pszemraj/pythia-31m-simplewiki-scratch-bf16 on the Open LLM Leaderboard.
The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_pszemraj__pythia-31m-simplewiki-scratch-bf16.details_OpenBuddy__openbuddy-codellama2-34b-v11.1-bf16
Dataset Card for Evaluation run of OpenBuddy/openbuddy-codellama2-34b-v11.1-bf16
Dataset Summary
Dataset automatically created during the evaluation run of model OpenBuddy/openbuddy-codellama2-34b-v11.1-bf16 on the Open LLM Leaderboard.
The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 3 run(s). Each run can be found as a specific split in each configuration, the split being named using the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_OpenBuddy__openbuddy-codellama2-34b-v11.1-bf16.details_DevaMalla__llama7b_alpaca_1gpu_bf16
Dataset Card for Evaluation run of DevaMalla/llama7b_alpaca_1gpu_bf16
Dataset Summary
Dataset automatically created during the evaluation run of model DevaMalla/llama7b_alpaca_1gpu_bf16 on the Open LLM Leaderboard.
The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_DevaMalla__llama7b_alpaca_1gpu_bf16.details_CultriX__NeuralTrixlaser-bf16
Dataset Card for Evaluation run of CultriX/NeuralTrixlaser-bf16
Dataset automatically created during the evaluation run of model CultriX/NeuralTrixlaser-bf16 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_CultriX__NeuralTrixlaser-bf16.details_Edgerunners__yi-9b-may-ortho-baukit-13fail-3000total-bf16BF16kEval_FinEval_16k_fulleval__3args_ours-eval_rldetails_fblgit__una-cybertron-7b-v2-bf16
Dataset Card for Evaluation run of fblgit/una-cybertron-7b-v2-bf16
Dataset Summary
Dataset automatically created during the evaluation run of model fblgit/una-cybertron-7b-v2-bf16 on the Open LLM Leaderboard.
The dataset is composed of 63 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_fblgit__una-cybertron-7b-v2-bf16.lm-eval-results-CultriX-NeuralTrix-bf16-private
Dataset Card for Evaluation run of CultriX/NeuralTrix-bf16
Dataset automatically created during the evaluation run of model CultriX/NeuralTrix-bf16
The dataset is composed of 62 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An… See the full description on the dataset page: https://huggingface.co/datasets/nyu-dice-lab/lm-eval-results-CultriX-NeuralTrix-bf16-private.