Generative Reranking on Avito
0.7541AUCCongrats
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CongratsModel Category=Generative, Approach=Non-autoregressive, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.7541 | 0.7553 | |
| NAR4RecModel Category=Generative, Approach=Non-autoregressive, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.7234 | 0.7409 | |
| Seq2SlateModel Category=Generative, Approach=Autoregressive, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.7134 | 0.7225 | |
| PIERModel Category=Traditional, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.7109 | 0.7401 | |
| Edge-rerankModel Category=Traditional, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.6953 | 0.7203 | |
| PRMModel Category=Traditional, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.6881 | 0.738 | |
| DCNModel Category=Traditional, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.6623 | 0.7004 | |
| DNNModel Category=Traditional, GPU=Tesla T4 (16G), Batch Size=10242025.10 | 0.6614 | 0.692 |