datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
asolaria-record-231-canonical
asolaria-record-231-canonical
The photographic record of Jesse Daniel Brown, in his own numbering.
What is here
path
what it is
photos/
231 photographs, numbered 001–231, each keeping its original camera filename after the number
MAPPING.tsv
the canonical index: number, path, original filename, byte size, SHA-256
CHECKSUMS-231.sha256
machine-checkable form of the same, for sha256sum -c
OBSERVATION.md
the written observation document, 8,194 lines… See the full description on the dataset page: https://huggingface.co/datasets/Jessedbrown/asolaria-record-231-canonical.smk_canon_paintingsimagenet-100-LT-canonical-eval
Canonical Balanced Evaluation Splits for ImageNet-100-LT
This dataset contains deterministic, class-balanced evaluation holdouts carved
from the pinned train split of
flaitenberger/imagenet-100-LT.
Splits
Split
Rows
Per class
validation
100
1
test
100
1
The source has only one long-tailed split, and its rarest class contains three
images. Therefore v1 uses one image per class for validation, one per class for
test, and leaves at least one source… See the full description on the dataset page: https://huggingface.co/datasets/quinnlue/imagenet-100-LT-canonical-eval.my_first_lora_v1-datasetdima2000-dataset