CoolFace
13 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
01Ryukijano /repro-causal-jepa-learning-world-models-through-object-level-latent-masking-traces Agent traces Agent sessions published from a Trackio Logbook. tabularn<1K0 likes104 downloads2mo agoHugging Face02drvp /world_model_eval_logstabularn<1K0 likes30 downloads2mo agoHugging Face03ultrastar111 /sokoban_easy_v8_noncot_chunk_k10_world_model_20260622_perseg sokoban_easy_v8_noncot_chunk_k10_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (non-CoT baseline) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding).… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_noncot_chunk_k10_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes17 downloads2mo agoHugging Face04ultrastar111 /sokoban_easy_v8_noncot_chunk_k5_world_model_20260622_perseg sokoban_easy_v8_noncot_chunk_k5_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (non-CoT baseline) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding).… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_noncot_chunk_k5_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes16 downloads2mo agoHugging Face05ultrastar111 /sokoban_easy_v8_noncot_chunk_k3_world_model_20260622_perseg sokoban_easy_v8_noncot_chunk_k3_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (non-CoT baseline) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding).… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_noncot_chunk_k3_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes14 downloads2mo agoHugging Face06ultrastar111 /sokoban_easy_v8_cot_chunk_kinf_world_model_20260707_perseg sokoban_easy_v8_cot_chunk_kinf_world_model_20260707_perseg Sokoban action-conditioned visual world-model SFT data (CoT self-rollout) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding). See… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_cot_chunk_kinf_world_model_20260707_perseg.tabularreinforcement-learning100K<n<1M0 likes13 downloads2mo agoHugging Face07ultrastar111 /sokoban_easy_v8_cot_chunk_k1_world_model_20260622_perseg sokoban_easy_v8_cot_chunk_k1_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (CoT self-rollout) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding). See… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_cot_chunk_k1_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes9 downloads2mo agoHugging Face08drvp /world_model_evaltabularn<1K0 likes8 downloads4mo agoHugging Face09ultrastar111 /sokoban_easy_v8_cot_chunk_k3_world_model_20260622_perseg sokoban_easy_v8_cot_chunk_k3_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (CoT self-rollout) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding). See… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_cot_chunk_k3_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes7 downloads2mo agoHugging Face10ultrastar111 /sokoban_easy_v8_cot_chunk_k5_world_model_20260622_perseg sokoban_easy_v8_cot_chunk_k5_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (CoT self-rollout) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding). See… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_cot_chunk_k5_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes7 downloads2mo agoHugging Face11ultrastar111 /sokoban_easy_v8_cot_chunk_k10_world_model_20260622_perseg sokoban_easy_v8_cot_chunk_k10_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (CoT self-rollout) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding). See… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_cot_chunk_k10_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes6 downloads2mo agoHugging Face12LangAGI-Lab /world_model_for_wa_qatabular10K<n<100K0 likes5 downloads2y agoHugging Face13ultrastar111 /sokoban_easy_v8_noncot_chunk_k1_world_model_20260622_perseg sokoban_easy_v8_noncot_chunk_k1_world_model_20260622_perseg Sokoban action-conditioned visual world-model SFT data (non-CoT baseline) for the BAGEL-7B-MoT VLM-Gym feedback-interval study. Format: gzipped JSONL shards under training/, one packed row = one episode. Frames are base64 JPEG (q95). Per-segment CoT layout: <think> per-step imagined frame (MSE target) </think> then the committed action chunk; between chunks a loss-0 "Action executed." + real frame (GT re-grounding).… See the full description on the dataset page: https://huggingface.co/datasets/ultrastar111/sokoban_easy_v8_noncot_chunk_k1_world_model_20260622_perseg.tabularreinforcement-learning100K<n<1M0 likes5 downloads2mo agoHugging Face

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