Loss Functions

The training objectives that shape every model — regression, classification, distribution-matching, and contrastive.

1. 1. ○ Implement Mean Squared Error
2. 2. ○ Implement Binary Cross-Entropy Loss
3. 3. ○ Implement Cross-Entropy Loss
4. 4. ○ Implement KL Divergence
5. 5. ○ Label Smoothing
6. 6. ○ Focal Loss
7. 7. ○ Triplet Loss
8. 8. ○ Contrastive Loss (InfoNCE)