Add model card

README.md

@@ -84,11 +84,11 @@ Trained on wav2vec2 multi-layer embeddings covering 9 languages:
 ## Performance
 
 **Test Set Results (16,709 samples):**
-- **Overall Accuracy:** 82.28%
-- **Macro F1-Score:** 80.45%
+- **Overall Accuracy:** 86.25%
+- **Macro F1-Score:** 84.60%
 - **Weighted F1-Score:** 86.04%
 - **Classes:** negative, neutral, positive
-- **Epochs Trained:** 79
+- **Epochs Trained:** 111
 - **Training Samples:** 83,545
 
 ### Per-Class Performance
@@ -136,7 +136,7 @@ Hyperparameters:
   Optimizer: AdamW (lr=2e-4, weight_decay=0.01)
   Loss: CrossEntropyLoss (label_smoothing=0.1, class-weighted)
   LR Schedule: Cosine annealing + 5% warmup
-  Batch Size: 32, Epochs: 79
+  Batch Size: 32, Epochs: 111
   Regularization: Gradient clipping (max_norm=1.0)
 ```
 
@@ -318,5 +318,5 @@ If you use this model, please cite:
 
 ---
 
-**Uploaded:** 2025-11-14 14:31:57 UTC  
+**Uploaded:** 2025-11-14 15:56:29 UTC  
 **Version:** 3.0 (Transformer + Multilingual + Multi-Layer Features)