Sentiment Classification on Urdu Sentiment Classification Dataset
77.6AccuracyLLaMA 3.2
Evaluation results
| Method | Links | |
|---|---|---|
| LLaMA 3.2Parameters=1B, Shot-count=k=5, Averaging=5 runs2026.01 | 77.6 | |
| UrduLMParameters=100M, Shot-count=k=5, Averaging=5 runs2026.01 | 66.6 | |
| Gemma 2Parameters=2B, Shot-count=k=5, Averaging=5 runs2026.01 | 62.6 | |
| Qwen 2.5Parameters=1.5B, Shot-count=k=5, Averaging=5 runs2026.01 | 50.2 | |
| LLaMA 3.2Parameters=3B, Shot-count=k=5, Averaging=5 runs2026.01 | 50.2 | |
| Qwen 2.5Parameters=3B, Shot-count=k=5, Averaging=5 runs2026.01 | 50.02 |