panther
Datasets
All datasets matching “panther”pantherpanthera-ik-three-block-stack-30hz
Panthera IK three-block stacking
1,000 successful scripted IK demonstrations; 862,766
training frames at 30 Hz. This is the three-block Panthera MuJoCo task:
red onto green, then blue onto red. The stored task prompt is
stack the three colored cubes.
LeRobot v3.0 format, exported using LeRobot 0.4.4. Both 256×256 RGB cameras
(observation.images.shoulder, observation.images.wrist) are embedded JPEGs
in Parquet. No video decoder, raw-data conversion, or MuJoCo rendering is needed… See the full description on the dataset page: https://huggingface.co/datasets/FoxNerdSaysMoo/panthera-ik-three-block-stack-30hz.pantherpantherpanther-seg-a-original-archives
PANTHER and SEG-A original download archives
Private backup of five original ZIP archives covering three tasks: PANTHER Task 1, PANTHER Task 2, and SEG-A (Dongyang, KiTS, Rider).
panther-seg-a-original-archives.tar contains the original ZIP files, with task directories preserved, plus the manifest, original SHA-256 checksums, and PANTHER source README. PANTHER is stored once in panther/ (two distinct task ZIPs); SEG-A ZIPs are in seg_a/. ZIP bytes are unchanged. Prepared… See the full description on the dataset page: https://huggingface.co/datasets/Saint-lsy/panther-seg-a-original-archives.Panther-dataset_v1
Dataset Details
This dataset is a modified version of Anthropic/hh-rlhf
This dataset is used in fine tuning Panther - an state of the art LLM funtuned on llama-7b pretrained model.
A very small portion i.e. 5.3% of prompts and responses were taken from this dataset to finetune and train Panther
Dataset Details
Dataset Structure
Train
Train rows : 377k
Validation
Validation rows : 20.3k
Dataset Format
input… See the full description on the dataset page: https://huggingface.co/datasets/Rardilit/Panther-dataset_v1.