System Latency on MS MARCO (test)
0.55Latency P50 (s)TEI RadixMLP
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TEI RadixMLPModel=Qwen3-0.6B2026.01 | 0.55 | 0.6 | |
| vLLMModel=Qwen3-0.6B, version=v0.13.02026.01 | 0.71 | 0.79 | |
| TEIModel=Qwen3-0.6B2026.01 | 0.78 | 0.85 | |
| vLLMModel=Qwen3-4B, version=v0.13.02026.01 | 2.22 | 2.51 | |
| TEI RadixMLPModel=Qwen3-4B2026.01 | 2.39 | 2.62 | |
| vLLMModel=Qwen3-8B, version=v0.13.02026.01 | 3.4 | 3.84 | |
| TEIModel=Qwen3-4B2026.01 | 3.65 | 3.98 | |
| TEI RadixMLPModel=Qwen3-8B2026.01 | 3.66 | 4 | |
| TEIModel=Qwen3-8B2026.01 | 5.75 | 6.27 |