Dataset Viewer (First 5GB)
Search is not available for this dataset
x
array 3D | y
int16 0
255
|
|---|---|
[[[0.151123046875,0.8310546875,0.71240234375,0.6748046875,0.845703125,0.25830078125,0.1878662109375,(...TRUNCATED) | 71 |
[[[0.028778076171875,0.475830078125,0.5498046875,0.87060546875,0.15869140625,0.3642578125,0.41381835(...TRUNCATED) | 203 |
[[[0.1488037109375,0.82275390625,0.9541015625,0.39892578125,0.57666015625,0.366455078125,0.102478027(...TRUNCATED) | 189 |
[[[0.83251953125,0.45751953125,0.6875,0.079345703125,0.53955078125,0.935546875,0.211669921875,0.4680(...TRUNCATED) | 77 |
[[[0.54150390625,0.488525390625,0.1326904296875,0.53857421875,0.48095703125,0.17041015625,0.17651367(...TRUNCATED) | 4 |
[[[0.418701171875,0.490966796875,0.4609375,0.896484375,0.65771484375,0.9140625,0.2352294921875,0.687(...TRUNCATED) | 180 |
[[[0.697265625,0.114990234375,0.93408203125,0.2249755859375,0.407470703125,0.57666015625,0.228027343(...TRUNCATED) | 93 |
[[[0.74951171875,0.7001953125,0.4033203125,0.1153564453125,0.4521484375,0.50244140625,0.221313476562(...TRUNCATED) | 102 |
[[[0.6044921875,0.8974609375,0.0000553131103515625,0.352783203125,0.68896484375,0.17431640625,0.0679(...TRUNCATED) | 122 |
[[[0.5966796875,0.75048828125,0.5263671875,0.5498046875,0.6669921875,0.353759765625,0.86572265625,0.(...TRUNCATED) | 47 |
End of preview. Expand
in Data Studio
YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/datasets-cards)
FranceCrops Float16 Dataset
Memory-efficient version of the FranceCrops dataset stored in float16 format.
Usage
from datasets import load_dataset
# Load dataset
dataset = load_dataset("saget-antoine/francecrops-float16")
# Load with specific format (recommended for training)
# As float32
dataset = dataset.with_format('torch', dtype=torch.float32)
# Or as float16 for memory efficiency
dataset = dataset.with_format('torch', dtype=torch.float16)
# Or as bfloat16 for TPU compatibility
dataset = dataset.with_format('torch', dtype=torch.bfloat16)
Memory Usage
Using float16 format reduces the memory footprint by approximately 50% compared to float32, while maintaining sufficient precision for most agricultural monitoring tasks.
Citation
If you use this dataset, please cite the original FranceCrops paper.
- Downloads last month
- 9
Each sample contains Sentinel-2 time series data (shape: 100, 60, 12) and corresponding crop type labels
Each sample contains Sentinel-2 time series data (shape: 100, 60, 12) and corresponding crop type labels
Features stored in float16 format for reduced memory usage
Features stored in float16 format for reduced memory usage
Labels stored as int16
Labels stored as int16
Size of the auto-converted Parquet files (First 5GB):
3.9 GB
Number of rows (First 5GB):
30,688
Estimated number of rows:
198,336