philipphager
commited on
Commit
•
e707f58
1
Parent(s):
44584ab
Update README.md
Browse files
README.md
CHANGED
@@ -16,17 +16,17 @@ metrics:
|
|
16 |
# Naive Listwise MonoBERT trained on Baidu-ULTR
|
17 |
A flax-based MonoBERT cross encoder trained on the [Baidu-ULTR](https://arxiv.org/abs/2207.03051) dataset with a **listwise softmax cross-entropy loss on clicks**. The loss is called "naive" as we use user clicks as a signal of relevance without any additional position bias correction. For more info, [read our paper](https://arxiv.org/abs/2404.02543) and [find the code for this model here](https://github.com/philipphager/baidu-bert-model).
|
18 |
|
19 |
-
## Test Results on Baidu-ULTR
|
20 |
-
|
21 |
-
| Model | Log-likelihood | DCG@1 | DCG@3 | DCG@5 | DCG@10 | nDCG@10 | MRR@10 |
|
22 |
-
|---------------------|----------------|-------|-------|-------|--------|---------|--------|
|
23 |
-
| Pointwise Naive | 0.227 | 1.641 | 3.462 | 4.752 | 7.251 | 0.357 | 0.609 |
|
24 |
-
| Pointwise Two-Tower | 0.218 | 1.629 | 3.471 | 4.822 | 7.456 | 0.367 | 0.607 |
|
25 |
-
| Pointwise IPS | 0.222 | 1.295 | 2.811 | 3.977 | 6.296 | 0.307 | 0.534 |
|
26 |
-
| Listwise Naive | - | 1.947 | 4.108 | 5.614 | 8.478 | 0.405 | 0.639 |
|
27 |
-
| Listwise IPS | - | 1.671 | 3.530 | 4.873 | 7.450 | 0.361 | 0.603 |
|
28 |
-
| Listwise DLA | - | 1.796 | 3.730 | 5.125 | 7.802 | 0.377 | 0.615 |
|
29 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 |
|
31 |
|
32 |
## Usage
|
|
|
16 |
# Naive Listwise MonoBERT trained on Baidu-ULTR
|
17 |
A flax-based MonoBERT cross encoder trained on the [Baidu-ULTR](https://arxiv.org/abs/2207.03051) dataset with a **listwise softmax cross-entropy loss on clicks**. The loss is called "naive" as we use user clicks as a signal of relevance without any additional position bias correction. For more info, [read our paper](https://arxiv.org/abs/2404.02543) and [find the code for this model here](https://github.com/philipphager/baidu-bert-model).
|
18 |
|
19 |
+
## Test Results on Baidu-ULTR
|
20 |
+
Ranking performance is measured in DCG, nDCG, and MRR on expert annotations (6,985 queries). Click prediction performance is measured in log-likelihood on one test partition of user clicks (49,495 queries).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
21 |
|
22 |
+
| Model | Log-likelihood | DCG@1 | DCG@3 | DCG@5 | DCG@10 | nDCG@10 | MRR@10 |
|
23 |
+
|------------------------------------------------------------------------------------------------|----------------|-------|-------|-------|--------|---------|--------|
|
24 |
+
| [Pointwise Naive](https://huggingface.co/philipphager/baidu-ultr_uva-bert_naive-pointwise) | 0.227 | 1.641 | 3.462 | 4.752 | 7.251 | 0.357 | 0.609 |
|
25 |
+
| [Pointwise Two-Tower](https://huggingface.co/philipphager/baidu-ultr_uva-bert_twotower) | 0.218 | 1.629 | 3.471 | 4.822 | 7.456 | 0.367 | 0.607 |
|
26 |
+
| [Pointwise IPS](https://huggingface.co/philipphager/baidu-ultr_uva-bert_ips-pointwise) | 0.222 | 1.295 | 2.811 | 3.977 | 6.296 | 0.307 | 0.534 |
|
27 |
+
| [Listwise Naive](https://huggingface.co/philipphager/baidu-ultr_uva-bert_naive-listwise) | - | 1.947 | 4.108 | 5.614 | 8.478 | 0.405 | 0.639 |
|
28 |
+
| [Listwise IPS](https://huggingface.co/philipphager/baidu-ultr_uva-bert_ips-listwise) | - | 1.671 | 3.530 | 4.873 | 7.450 | 0.361 | 0.603 |
|
29 |
+
| [Listwise DLA](https://huggingface.co/philipphager/baidu-ultr_uva-bert_dla) | - | 1.796 | 3.730 | 5.125 | 7.802 | 0.377 | 0.615 |
|
30 |
|
31 |
|
32 |
## Usage
|