Bonkh/lid-stack-model-gemma_3_27b-gemma_3_300m

This is the expand version of dleemiller/WordLlamaDetect model. Stacking two wordllama-based models to enhance the performace

• Support languages: 148

Training data (740k samples)
        │
        ▼
┌───────────────────────────────────┐
│         Phase 1: Base Models      │
│                                   │
│  ┌─────────────┐  ┌─────────────┐ │
│  │LID Model 01 │  │ LID Model 02│ │
│  └──────┬──────┘  └──────┬──────┘ │
└─────────┼────────────────┼────────┘
          │  train each    │
          │  independently │
          ▼                ▼
     lid_models[0]    lid_models[1]
          │                │
          └───────┬────────┘
                  │
                  ▼
          collect_preds() → X: (N, 2*148) = (N, 296)

model1 logits      model2 logits
  (N, 148)    cat    (N, 148)
      └────────┬──────────┘
               ▼
           (N, 296)
               │
       Linear(296 → 148)           ← 296*148 = 43,808 params trained
               │
               ▼
          (N, 148) → CrossEntropy(y)

Evaluation results on Flores +

Pair	Num Languages	Accuracy	F1 Macro	Metric per Base Model
gemma3_27b + gemma_300m	148	0.9307	0.9303	gemma3_27b: Acc: 0.9147, F1: 0.9149 gemma_300m: Acc: 0.9087, F1: 0.9078