Migrate model card from transformers-repo
Browse filesRead announcement at https://discuss.huggingface.co/t/announcement-all-model-cards-will-be-migrated-to-hf-co-model-repos/2755
Original file history: https://github.com/huggingface/transformers/commits/master/model_cards/mrm8488/RuPERTa-base-finetuned-pos/README.md
README.md
ADDED
|
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: es
|
| 3 |
+
thumbnail:
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# RuPERTa-base (Spanish RoBERTa) + POS 🎃🏷
|
| 7 |
+
|
| 8 |
+
This model is a fine-tuned on [CONLL CORPORA](https://www.kaggle.com/nltkdata/conll-corpora) version of [RuPERTa-base](https://huggingface.co/mrm8488/RuPERTa-base) for **POS** downstream task.
|
| 9 |
+
|
| 10 |
+
## Details of the downstream task (POS) - Dataset
|
| 11 |
+
|
| 12 |
+
- [Dataset: CONLL Corpora ES](https://www.kaggle.com/nltkdata/conll-corpora) 📚
|
| 13 |
+
|
| 14 |
+
| Dataset | # Examples |
|
| 15 |
+
| ---------------------- | ----- |
|
| 16 |
+
| Train | 445 K |
|
| 17 |
+
| Dev | 55 K |
|
| 18 |
+
|
| 19 |
+
- [Fine-tune on NER script provided by Huggingface](https://github.com/huggingface/transformers/blob/master/examples/token-classification/run_ner_old.py)
|
| 20 |
+
|
| 21 |
+
- Labels covered:
|
| 22 |
+
|
| 23 |
+
```
|
| 24 |
+
ADJ
|
| 25 |
+
ADP
|
| 26 |
+
ADV
|
| 27 |
+
AUX
|
| 28 |
+
CCONJ
|
| 29 |
+
DET
|
| 30 |
+
INTJ
|
| 31 |
+
NOUN
|
| 32 |
+
NUM
|
| 33 |
+
PART
|
| 34 |
+
PRON
|
| 35 |
+
PROPN
|
| 36 |
+
PUNCT
|
| 37 |
+
SCONJ
|
| 38 |
+
SYM
|
| 39 |
+
VERB
|
| 40 |
+
```
|
| 41 |
+
|
| 42 |
+
## Metrics on evaluation set 🧾
|
| 43 |
+
|
| 44 |
+
| Metric | # score |
|
| 45 |
+
| :------------------------------------------------------------------------------------: | :-------: |
|
| 46 |
+
| F1 | **97.39**
|
| 47 |
+
| Precision | **97.47** |
|
| 48 |
+
| Recall | **9732** |
|
| 49 |
+
|
| 50 |
+
## Model in action 🔨
|
| 51 |
+
|
| 52 |
+
|
| 53 |
+
Example of usage
|
| 54 |
+
|
| 55 |
+
```python
|
| 56 |
+
import torch
|
| 57 |
+
from transformers import AutoModelForTokenClassification, AutoTokenizer
|
| 58 |
+
|
| 59 |
+
tokenizer = AutoTokenizer.from_pretrained('mrm8488/RuPERTa-base-finetuned-pos')
|
| 60 |
+
model = AutoModelForTokenClassification.from_pretrained('mrm8488/RuPERTa-base-finetuned-pos')
|
| 61 |
+
|
| 62 |
+
id2label = {
|
| 63 |
+
"0": "O",
|
| 64 |
+
"1": "ADJ",
|
| 65 |
+
"2": "ADP",
|
| 66 |
+
"3": "ADV",
|
| 67 |
+
"4": "AUX",
|
| 68 |
+
"5": "CCONJ",
|
| 69 |
+
"6": "DET",
|
| 70 |
+
"7": "INTJ",
|
| 71 |
+
"8": "NOUN",
|
| 72 |
+
"9": "NUM",
|
| 73 |
+
"10": "PART",
|
| 74 |
+
"11": "PRON",
|
| 75 |
+
"12": "PROPN",
|
| 76 |
+
"13": "PUNCT",
|
| 77 |
+
"14": "SCONJ",
|
| 78 |
+
"15": "SYM",
|
| 79 |
+
"16": "VERB"
|
| 80 |
+
}
|
| 81 |
+
|
| 82 |
+
text ="Mis amigos están pensando viajar a Londres este verano."
|
| 83 |
+
input_ids = torch.tensor(tokenizer.encode(text)).unsqueeze(0)
|
| 84 |
+
|
| 85 |
+
outputs = model(input_ids)
|
| 86 |
+
last_hidden_states = outputs[0]
|
| 87 |
+
|
| 88 |
+
for m in last_hidden_states:
|
| 89 |
+
for index, n in enumerate(m):
|
| 90 |
+
if(index > 0 and index <= len(text.split(" "))):
|
| 91 |
+
print(text.split(" ")[index-1] + ": " + id2label[str(torch.argmax(n).item())])
|
| 92 |
+
|
| 93 |
+
'''
|
| 94 |
+
Output:
|
| 95 |
+
--------
|
| 96 |
+
Mis: NUM
|
| 97 |
+
amigos: PRON
|
| 98 |
+
están: AUX
|
| 99 |
+
pensando: ADV
|
| 100 |
+
viajar: VERB
|
| 101 |
+
a: ADP
|
| 102 |
+
Londres: PROPN
|
| 103 |
+
este: DET
|
| 104 |
+
verano..: NOUN
|
| 105 |
+
'''
|
| 106 |
+
```
|
| 107 |
+
Yeah! Not too bad 🎉
|
| 108 |
+
|
| 109 |
+
> Created by [Manuel Romero/@mrm8488](https://twitter.com/mrm8488) | [LinkedIn](https://www.linkedin.com/in/manuel-romero-cs/)
|
| 110 |
+
|
| 111 |
+
> Made with <span style="color: #e25555;">♥</span> in Spain
|