Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
datasets:
|
| 3 |
+
- upb-nlp/same_topic_articles
|
| 4 |
+
language:
|
| 5 |
+
- ro
|
| 6 |
+
- en
|
| 7 |
+
base_model:
|
| 8 |
+
- FacebookAI/xlm-roberta-large
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
## How to Get Started with the Model
|
| 12 |
+
|
| 13 |
+
Use the code below to get started with the model.
|
| 14 |
+
|
| 15 |
+
```python
|
| 16 |
+
import torch
|
| 17 |
+
from transformers import AutoTokenizer, XLMRobertaForSequenceClassification
|
| 18 |
+
|
| 19 |
+
MODEL_PATH = "upb-nlp/xlm_roberta_large_article_same_topic_classification"
|
| 20 |
+
|
| 21 |
+
tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
|
| 22 |
+
model = XLMRobertaForSequenceClassification.from_pretrained(MODEL_PATH, num_labels=2).to('cuda')
|
| 23 |
+
model.eval()
|
| 24 |
+
|
| 25 |
+
t1 = "Article title. Article body."
|
| 26 |
+
t2 = "Article title. Article body."
|
| 27 |
+
|
| 28 |
+
inputs = tokenizer(
|
| 29 |
+
t1,
|
| 30 |
+
t2,
|
| 31 |
+
return_tensors="pt",
|
| 32 |
+
truncation=True,
|
| 33 |
+
padding='max_length',
|
| 34 |
+
max_length=512
|
| 35 |
+
).to('cuda')
|
| 36 |
+
|
| 37 |
+
# Generate prediction
|
| 38 |
+
with torch.no_grad():
|
| 39 |
+
outputs = model(**inputs)
|
| 40 |
+
logits = outputs.logits
|
| 41 |
+
predicted_class = torch.argmax(logits, dim=1).item()
|
| 42 |
+
|
| 43 |
+
print(predicted_class)
|
| 44 |
+
```
|