Speculative Decoding Inference on PubMedQA
182.24Throughput (tokens/s)EvoSpec
Evaluation Results
| Method | Links | |
|---|---|---|
| EvoSpecStatic Vocabulary Size=32k, Dynamic Expansion Size=2562026.04 | 182.24 | |
| EvoSpecStatic Vocabulary Size=32k, Dynamic Expansion Size=5122026.04 | 181.31 | |
| EvoSpecStatic Vocabulary Size=32k, Dynamic Expansion Size=10242026.04 | 180.98 | |
| CORAL2026.04 | 178.11 | |
| EvoSpecStatic Vocabulary Size=16k, Dynamic Expansion Size=5122026.04 | 171.53 | |
| EvoSpecStatic Vocabulary Size=16k, Dynamic Expansion Size=2562026.04 | 169.98 | |
| EvoSpecStatic Vocabulary Size=16k, Dynamic Expansion Size=10242026.04 | 168.37 | |
| FR-SpecStatic Vocabulary Size=32k2026.04 | 160.56 | |
| EAGLE-32026.04 | 160.41 | |
| EvoSpecStatic Vocabulary Size=8k, Dynamic Expansion Size=10242026.04 | 157.4 | |
| EvoSpecStatic Vocabulary Size=8k, Dynamic Expansion Size=5122026.04 | 154.99 | |
| EvoSpecStatic Vocabulary Size=8k, Dynamic Expansion Size=2562026.04 | 154.27 |