Automatic Speech Recognition on Slavic Global (test)
6.7CERembedding-based H-Softmax
Evaluation results
| Method | Links | |
|---|---|---|
| embedding-based H-SoftmaxEmbedding Source=XLM-base, Clustering Method=2-medoids, Distance Metric=Standardized Euclidean2025.01 | 6.7 | |
| embedding-based H-SoftmaxEmbedding Source=XLM-large, Clustering Method=2-medoids, Distance Metric=Standardized Euclidean2025.01 | 6.7 | |
| embedding-based H-SoftmaxEmbedding Source=LABSE, Clustering Method=Median, Distance Metric=Euclidean2025.01 | 6.8 | |
| Huffman H-SoftmaxOutput Strategy=H-Softmax, Clustering=Huffman2025.01 | 6.9 | |
| embedding-based H-SoftmaxEmbedding Source=Mono-Map, Clustering Method=Average, Distance Metric=City Block2025.01 | 7.1 | |
| embedding-based H-SoftmaxEmbedding Source=XLM-XL, Clustering Method=Weighted, Distance Metric=Correlation2025.01 | 7.2 | |
| SoftmaxOutput Strategy=Softmax2025.01 | 8.4 |