README.md

---
inference: false
language:
- bg
license: mit
datasets:
- oscar
- chitanka
- wikipedia
tags:
- torch
---

# ROBERTA BASE (cased) trained on private Bulgarian sentiment-analysis dataset
This is a Multilingual Roberta model. 

This model is cased: it does make a difference between bulgarian and Bulgarian.

### How to use

Here is how to use this model in PyTorch:

```python
>>> import torch
>>> from transformers import AutoModel, AutoTokenizer
>>> 
>>> model_id = "rmihaylov/roberta-base-sentiment-bg"
>>> model = AutoModel.from_pretrained(model_id, trust_remote_code=True)
>>> tokenizer = AutoTokenizer.from_pretrained(model_id)
>>>
>>> inputs = tokenizer.batch_encode_plus(['Това е умно.', 'Това е тъпо.'], return_tensors='pt')
>>> outputs = model(**inputs)
>>> torch.softmax(outputs, dim=1).tolist()

[[0.0004746630438603461, 0.9995253086090088],
 [0.9986956715583801, 0.0013043134240433574]]
```