Add a 0.74 confidence threshold so speculative drafting stops early once the draft model's predicted token probability drops below it, favoring shorter, higher-acceptance draft sequences. |
||
|---|---|---|
| .. | ||
| amd | ||
| attic | ||
| halo | ||
| mx | ||
| nixtee1 | ||
| sgx | ||
| t15 | ||
| x1 | ||
Add a 0.74 confidence threshold so speculative drafting stops early once the draft model's predicted token probability drops below it, favoring shorter, higher-acceptance draft sequences. |
||
|---|---|---|
| .. | ||
| amd | ||
| attic | ||
| halo | ||
| mx | ||
| nixtee1 | ||
| sgx | ||
| t15 | ||
| x1 | ||