f4beaac559e00a3676d942dc7e8fea69efc01cfe,catalyst/metrics/avg_precision.py,,mean_avg_precision,#Any#Any#Any#,74
Before Change
The map score for every k.
size: [len(top_k), 1]
map_k_tuple = tuple(
torch.mean(avg_precision(outputs, targets, k)).item() for k in top_k
)
return map_k_tuple
__all__ = ["mean_avg_precision", "avg_precision"]
After Change
The map score for every k.
size: len(top_k)
results = []
for k in topk:
k = min(outputs.size(1), k)
results.append(torch.mean(avg_precision(outputs, targets)[:k]))
return results
__all__ = ["mean_avg_precision", "avg_precision"]
In pattern: SUPERPATTERN
Frequency: 3
Non-data size: 9
Instances
Project Name: catalyst-team/catalyst
Commit Name: f4beaac559e00a3676d942dc7e8fea69efc01cfe
Time: 2020-11-30
Author: raveforlive@gmail.com
File Name: catalyst/metrics/avg_precision.py
Class Name:
Method Name: mean_avg_precision
Project Name: IndicoDataSolutions/finetune
Commit Name: 2ad9e8af614767e645c4358ae1e2444f02d4573b
Time: 2018-08-22
Author: madison@indico.io
File Name: finetune/sequence_labeling.py
Class Name: SequenceLabeler
Method Name: predict