PII detection on PUPA
68.48F1 ScoreFine-tuning
Evaluation results
| Method | Links | |
|---|---|---|
| Fine-tuningTrain Tokens (Base Model)=190881 / 7086, Inference Tokens (Base Model)=19263 / 6112026.04 | 68.48 | |
| DSPy MIPROv2Train Tokens (Base Model)=1490233 / 134750, Inference Tokens (Base Model)=52409 / 14802026.04 | 64.91 | |
| AIRTrain Tokens (Embed Model)=90088 / -, Train Tokens (Base Model)=329236 / 2451, Train Tokens (Reflection Model)=55382 / 57279, Inference Tokens (Base Model)=252543 / 6212026.04 | 59.32 | |
| DSPy GEPATrain Tokens (Base Model)=1321431 / 229379, Train Tokens (Reflection Model)=24659 / 13374, Inference Tokens (Base Model)=71669 / 15742026.04 | 59.29 | |
| TextGradTrain Tokens (Base Model)=1298401 / 14733, Train Tokens (Reflection Model)=161048 / 177955, Inference Tokens (Base Model)=123063 / 9862026.04 | 55.54 | |
| DSPy BootstrapFewShot (#3)Train Tokens (Base Model)=16281 / 237, Inference Tokens (Base Model)=37889 / 14752026.04 | 44.23 | |
| KNNInference Tokens (Base Model)=34052 / 29652026.04 | 35.08 | |
| Initial promptInference Tokens (Base Model)=19263 / 26832026.04 | 17.58 |