---
license: mit
---

This classification model is based on [sberbank-ai/ruRoberta-large](https://huggingface.co/sberbank-ai/ruRoberta-large).
The model should be used to produce relevance and specificity of the last message in the context of a dialog.

It is pretrained on corpus of dialog data from social networks and finetuned on [tinkoff-ai/context_similarity](https://huggingface.co/tinkoff-ai/context_similarity). 
The performance of the model on validation split [tinkoff-ai/context_similarity](https://huggingface.co/tinkoff-ai/context_similarity) (with the best thresholds for validation samples):

<table>
    <thead>
        <tr>
            <td colspan="2">relevance</td>
            <td colspan="2">specificity</td>
        </tr>
    </thead>
    <tbody>
        <tr>
            <td>f0.5</td>
            <td>roc-auc</td>
            <td>f0.5</td>
            <td>roc-auc</td>
        </tr>
        <tr>
            <td>0.86</td>
            <td>0.83</td>
            <td>0.85</td>
            <td>0.86</td>
        </tr>
    </tbody>
</table>

The model can be loaded as follows:

```python
# pip install transformers
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained("tinkoff-ai/context_similarity")
model = AutoModel.from_pretrained("tinkoff-ai/context_similarity")
# model.cuda()
```