Datasets:
Tasks:
Image Segmentation
Modalities:
Image
Formats:
parquet
Sub-tasks:
semantic-segmentation
Size:
10K - 100K
Train/test split
#4
by
ckotait
- opened
Hello, Do you know where I can find the train/test splits of this dataset? Thanks!
Hi, I don't recall there being a train and test split when I found the dataset, but for my own training I split it like this:
ds = load_dataset("mattmdjaga/human_parsing_dataset", split="train[:100%]", num_proc=8)
split_ratio: float = 0.006
ds_split = ds.train_test_split(test_size=split_ratio)
train_ds = ds_split["train"]
test_ds = ds_split["test"]