LISBET
Collection
LISBET Is a Social Behavior Transformer • 3 items • Updated
Result on Task 1 - Frame classification (test set)
WARNING: Ignoring index 3 in macro metrics. A bug in torchmetrics may cause incorrect macro F1 and accuracy (see https://github.com/Lightning-AI/torchmetrics/issues/2441). Please consider validating your results agaist the per-class metrics.
Evaluation Summary Mode: multiclass Macro F1: 0.657 Macro Accuracy: 0.675 Per-class metrics: Class 0: F1=0.728, Precision=0.596, Recall=0.935 Class 1: F1=0.829, Precision=0.834, Recall=0.824 Class 2: F1=0.904, Precision=0.869, Recall=0.941 Class 3: F1=0.943, Precision=0.972, Recall=0.916
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support