Update README.md
Browse files
README.md
CHANGED
|
@@ -8,18 +8,6 @@ pipeline_tag: sentence-similarity
|
|
| 8 |
# SimCSE(sup)
|
| 9 |
|
| 10 |
|
| 11 |
-
## Model List
|
| 12 |
-
The evaluation dataset is in Chinese, and we used the same language model **RoBERTa base** on different methods.
|
| 13 |
-
| Model | STS-B(w-avg) | ATEC | BQ | LCQMC | PAWSX | Avg. |
|
| 14 |
-
|:-----------------------:|:------------:|:-----------:|:----------|:-------------|:------------:|:----------:|
|
| 15 |
-
| BERT-Whitening | 65.27| -| -| -| -| -|
|
| 16 |
-
| SimBERT | 70.01| -| -| -| -| -|
|
| 17 |
-
| SBERT-Whitening | 71.75| -| -| -| -| -|
|
| 18 |
-
| [BAAI/bge-base-zh](https://huggingface.co/BAAI/bge-base-zh) | 78.61| -| -| -| -| -|
|
| 19 |
-
| [hellonlp/simcse-base-zh(sup)](https://huggingface.co/hellonlp/simcse-roberta-base-zh) | **80.96**| -| -| -| -| -|
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
## Data List
|
| 24 |
The following datasets are all in Chinese.
|
| 25 |
| Data | size(train) | size(valid) | size(test) |
|
|
@@ -35,6 +23,28 @@ The following datasets are all in Chinese.
|
|
| 35 |
|
| 36 |
|
| 37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
## Uses
|
| 39 |
You can use our model for encoding sentences into embeddings
|
| 40 |
```python
|
|
|
|
| 8 |
# SimCSE(sup)
|
| 9 |
|
| 10 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
## Data List
|
| 12 |
The following datasets are all in Chinese.
|
| 13 |
| Data | size(train) | size(valid) | size(test) |
|
|
|
|
| 23 |
|
| 24 |
|
| 25 |
|
| 26 |
+
## Model List
|
| 27 |
+
The evaluation dataset is in Chinese, and we used the same language model **RoBERTa base** on different methods. In addition, considering that the test set of some datasets is small, which may lead to a large deviation in evaluation accuracy, the evaluation data here uses train, valid and test at the same time, and the final evaluation result adopts the **weighted average (w-avg)** method.
|
| 28 |
+
|
| 29 |
+
| Model | STS-B(w-avg) | ATEC | BQ | LCQMC | PAWSX | Avg. |
|
| 30 |
+
|:-----------------------:|:------------:|:-----------:|:----------|:----------|:----------:|:----------:|
|
| 31 |
+
| [BAAI/bge-large-zh](https://huggingface.co/BAAI/bge-large-zh) | 78.61| -| -| -| -| -|
|
| 32 |
+
| [BAAI/bge-large-zh-v1.5](https://huggingface.co/BAAI/bge-large-zh-v1.5) | 79.07| -| -| -| -| -|
|
| 33 |
+
| [hellonlp/simcse-large-zh](https://huggingface.co/hellonlp/simcse-roberta-large-zh) | 81.32| -| -| -| -| -|
|
| 34 |
+
|
| 35 |
+
|
| 36 |
+
| Model | STS-B(w-avg) | ATEC | BQ | LCQMC | PAWSX | Avg. |
|
| 37 |
+
|:-----------------------:|:------------:|:-----------:|:----------|:-------------|:------------:|:----------:|
|
| 38 |
+
| BERT-Whitening | 65.27| -| -| -| -| -|
|
| 39 |
+
| SimBERT | 70.01| -| -| -| -| -|
|
| 40 |
+
| SBERT-Whitening | 71.75| -| -| -| -| -|
|
| 41 |
+
| [BAAI/bge-base-zh](https://huggingface.co/BAAI/bge-base-zh) | 78.61| -| -| -| -| -|
|
| 42 |
+
| [hellonlp/simcse-base-zh(sup)](https://huggingface.co/hellonlp/simcse-roberta-base-zh) | **80.96**| -| -| -| -| -|
|
| 43 |
+
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
|
| 48 |
## Uses
|
| 49 |
You can use our model for encoding sentences into embeddings
|
| 50 |
```python
|