Update README.md
Browse files
README.md
CHANGED
|
@@ -101,7 +101,7 @@ tags:
|
|
| 101 |
- web-register
|
| 102 |
- genre
|
| 103 |
---
|
| 104 |
-
# Web register classification (
|
| 105 |
|
| 106 |
A web register classifier for texts in English, fine-tuned from [XLM-RoBERTa-large](https://huggingface.co/FacebookAI/xlm-roberta-large).
|
| 107 |
The model is trained with the [Corpus of Online Registers of English (CORE)](https://github.com/TurkuNLP/CORE-corpus) to classify documents based on the [CORE taxonomy](https://turkunlp.org/register-annotation-docs/).
|
|
@@ -226,7 +226,7 @@ Micro-averaged F1 scores and optimized prediction thresholds (test set):
|
|
| 226 |
|
| 227 |
| Language | F1 (All labels) | F1 (Main labels) | Threshold |
|
| 228 |
| -------- | --------------- | ---------------- | ----------|
|
| 229 |
-
| English | 0.74 | 0.
|
| 230 |
|
| 231 |
|
| 232 |
## Technical Specifications
|
|
|
|
| 101 |
- web-register
|
| 102 |
- genre
|
| 103 |
---
|
| 104 |
+
# Web register classification (English model)
|
| 105 |
|
| 106 |
A web register classifier for texts in English, fine-tuned from [XLM-RoBERTa-large](https://huggingface.co/FacebookAI/xlm-roberta-large).
|
| 107 |
The model is trained with the [Corpus of Online Registers of English (CORE)](https://github.com/TurkuNLP/CORE-corpus) to classify documents based on the [CORE taxonomy](https://turkunlp.org/register-annotation-docs/).
|
|
|
|
| 226 |
|
| 227 |
| Language | F1 (All labels) | F1 (Main labels) | Threshold |
|
| 228 |
| -------- | --------------- | ---------------- | ----------|
|
| 229 |
+
| English | 0.74 | 0.76 | 0.40 |
|
| 230 |
|
| 231 |
|
| 232 |
## Technical Specifications
|