Update README.md

README.md

@@ -1,23 +1,54 @@
 ---
-language: 
+language:
 - en
 - de
 - fr
-- zh
+- zh 
+- it
+- es
+- hi 
+- bn
+- ar
+- ru 
+- uk
+- pt
+- ur
+- id
+- ja
 - ne
-- multilingual
+- nl
+- tr
+- ca
+- bg
+- yue
+
 widget:
-- text: "In December 1903 in France the Royal Swedish Academy of Sciences awarded Pierre Curie, Marie Curie, and Henri Becquerel the Nobel Prize in Physics."
-- text: "Für Richard Phillips Feynman war es immer wichtig in New York, die unanschaulichen Gesetzmäßigkeiten der Quantenphysik Laien und Studenten nahezubringen und verständlich zu machen."
-- text: "My name is Julian and I live in Constance"
-- text: "Terence David John Pratchett est né le 28 avril 1948 à Beaconsfield dans le Buckinghamshire, en Angleterre."
-- text: "北京市，通称北京（汉语拼音：Běijīng；邮政式拼音：Peking），简称“京”，是中华人民共和国的首都及直辖市，是该国的政治、文化、科技、教育、军事和国际交往中心，是一座全球城市，是世界人口第三多的城市和人口最多的首都，具有重要的国际影响力，同時也是目前世界唯一的“双奥之城”，即唯一既主办过夏季"
-- text: "काठमाडौँ नेपालको सङ्घीय राजधानी र नेपालको सबैभन्दा बढी जनसङ्ख्या भएको सहर हो।"
+- text: >-
+    In December 1903 in France the Royal Swedish Academy of Sciences awarded
+    Pierre Curie, Marie Curie, and Henri Becquerel the Nobel Prize in Physics.
+- text: >-
+    Für Richard Phillips Feynman war es immer wichtig in New York, die
+    unanschaulichen Gesetzmäßigkeiten der Quantenphysik Laien und Studenten
+    nahezubringen und verständlich zu machen.
+- text: My name is Julian and I live in Constance.
+- text: >-
+    Terence David John Pratchett est né le 28 avril 1948 à Beaconsfield dans le
+    Buckinghamshire, en Angleterre.
+- text: >-
+    北京市，通称北京（汉语拼音：Běijīng；邮政式拼音：Peking），简称“京”，是中华人民共和国的首都及直辖市，是该国的政治、文化、科技、教育、军事和国际交往中心，是一座全球城市，是世界人口第三多的城市和人口最多的首都，具有重要的国际影响力，同時也是目前世界唯一的“双奥之城”，即唯一既主办过夏季
+
 tags:
 - roberta
+- ner
+- nlp
 license: mit
 datasets:
 - wikiann
+metrics:
+- f1
+- precision
+- accuracy
+- recall
 ---
 
 # RoBERTa for Multilingual Named Entity Recognition
@@ -30,70 +61,55 @@ This model detects entities by classifying every token according to the IOB form
 ['O', 'B-PER', 'I-PER', 'B-ORG', 'I-ORG', 'B-LOC', 'I-LOC']
 ```
 
-**Languages:**
-
-TBD
-
 ## Training data
 
-This mode was traind using a subset of all [wikiann](https://huggingface.co/datasets/wikiann) dataset.
-
-## Evaluation results
-
-This model achieves the following results (meassured using the validation portion of the [wikiann](https://huggingface.co/datasets/wikiann)):
+This model was fine-tuned on a portion of the [wikiann](https://huggingface.co/datasets/wikiann) dataset corresponding to the following languages:
 
 ```python
-{'LOC': {'f1': 0.8541617653978262,
-         'number': 42016,
-         'precision': 0.8444273885942878,
-         'recall': 0.8641231911652704},
- 'ORG': {'f1': 0.7504633739856393,
-         'number': 31226,
-         'precision': 0.7305394669011736,
-         'recall': 0.7715045154678793},
- 'PER': {'f1': 0.8639735635284596,
-         'number': 29647,
-         'precision': 0.863711444463172,
-         'recall': 0.8642358417377812},
- 'overall_accuracy': 0.926459490605155,
- 'overall_f1': 0.8250250567072849,
- 'overall_precision': 0.814290312198262,
- 'overall_recall': 0.8360466133405904}
-
+["en","de", "fr",
+ "zh", "it", "es",
+ "hi", "bn", "ar",
+ "ru", "uk", "pt",
+ "ur", "id", "ja",
+ "ne", "nl", "tr",
+ "ca", "bg", "zh-yue"]
 ```
 
-**Per Entity Type:**
-
-TBD
-
-**Per Language:**
-
-TBD
+The model was fine-tuned on 375.100 sentences in the training set, with a validation set of 173.100 examples. Performance metrics reported are based on additional 173.100 examples. The complete WikiANN dataset includes training examples for 282 languages and was constructed from Wikipedia. Training examples are extracted in an automated manner, exploiting entities mentioned in Wikipedia articles, often are formatted as hyperlinks to the source article. Provided NER tags are in the IOB2 format. Named entities are classified as location (LOC), person (PER), or organization (ORG). 
 
+## Evaluation results
 
-## About RoBERTa
-
-This model is a fine-tuned version of [XLM-RoBERTa](https://huggingface.co/xlm-roberta-large). The original model was pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. It was introduced in the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Conneau et al. and first released in [this repository](https://github.com/pytorch/fairseq/tree/master/examples/xlmr). 
-
-RoBERTa is a transformers model pretrained on a large corpus in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from those texts.
-
-More precisely, it was pretrained with the Masked language modeling (MLM) objective. Taking a sentence, the model randomly masks 15% of the words in the input then run the entire masked sentence through the model and has to predict the masked words. This is different from traditional recurrent neural networks (RNNs) that usually see the words one after the other, or from autoregressive models like GPT which internally mask the future tokens. It allows the model to learn a bidirectional representation of the sentence.
-
-This way, the model learns an inner representation of 100 languages that can then be used to extract features useful for downstream tasks: if you have a dataset of labeled sentences for instance, you can train a standard classifier using the features produced by the XLM-RoBERTa model as inputs.
+This model achieves the following results (meassured using the test split of the [wikiann](https://huggingface.co/datasets/wikiann) dataset):
 
-#### Limitations and bias
+```python
+{'LOC': {'f1': 0.9310524680196053,
+         'number': 545516,
+         'precision': 0.9230957726278464,
+         'recall': 0.9391475227124411},
+ 'ORG': {'f1': 0.884603763901478,
+         'number': 363324,
+         'precision': 0.8868243944134171,
+         'recall': 0.8823942266406843},
+ 'PER': {'f1': 0.939167449173159,
+         'number': 367750,
+         'precision': 0.934642687866253,
+         'recall': 0.9437362338545208},
+ 'overall_accuracy': 0.9588396024156357,
+ 'overall_f1': 0.9202625613733114,
+ 'overall_precision': 0.9162434124141294,
+ 'overall_recall': 0.9243171260937341}
 
-This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.  
+```
 
 ## Usage
 
-You can use this model by using the AutoTokenize and AutoModelForTokenClassification class:
+You can load this model by using the AutoTokenize and AutoModelForTokenClassification class:
 
 ```python
 from transformers import AutoTokenizer, AutoModelForTokenClassification 
 
-tokenizer = AutoTokenizer.from_pretrained("julian-schelb/roberta-ner-multilingual/", add_prefix_space=True)                          
-model = AutoModelForTokenClassification.from_pretrained("julian-schelb/roberta-ner-multilingual/")
+tokenizer = AutoTokenizer.from_pretrained("julian-schelb/roberta-ner-multilingual-wikiann/", add_prefix_space=True)                          
+model = AutoModelForTokenClassification.from_pretrained("julian-schelb/roberta-ner-multilingual-wikiann/")
 
 text = "In December 1903 in France the Royal Swedish Academy of Sciences awarded Pierre Curie, Marie Curie, and Henri Becquerel the Nobel Prize in Physics."
 
@@ -115,8 +131,24 @@ predicted_tokens_classes = [model.config.id2label[t.item()] for t in predicted_t
 predicted_tokens_classes
 ```
 
-### BibTeX entry and citation info
 
-```bibtex
-TBD
-```
+## About RoBERTa
+
+This model is a fine-tuned version of [XLM-RoBERTa](https://huggingface.co/xlm-roberta-large). The original model was pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. It was introduced in the paper [Unsupervised Cross-lingual Representation Learning at Scale](https://arxiv.org/abs/1911.02116) by Conneau et al. and first released in [this repository](https://github.com/pytorch/fairseq/tree/master/examples/xlmr). 
+
+RoBERTa is a transformers model pretrained on a large corpus in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from those texts.
+
+More precisely, it was pretrained with the Masked language modeling (MLM) objective. Taking a sentence, the model randomly masks 15% of the words in the input then run the entire masked sentence through the model and has to predict the masked words. This is different from traditional recurrent neural networks (RNNs) that usually see the words one after the other, or from autoregressive models like GPT which internally mask the future tokens. It allows the model to learn a bidirectional representation of the sentence.
+
+This way, the model learns an inner representation of 100 languages that can then be used to extract features useful for downstream tasks: if you have a dataset of labeled sentences for instance, you can train a standard classifier using the features produced by the XLM-RoBERTa model as inputs.
+
+#### Limitations and bias
+
+This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.  
+
+## Related Papers
+
+* Pan, X., Zhang, B., May, J., Nothman, J., Knight, K., & Ji, H. (2017). Cross-lingual Name Tagging and Linking for 282 Languages. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 1946–1958). Association for Computational Linguistics.
+* Rahimi, A., Li, Y., & Cohn, T. (2019). Massively Multilingual Transfer for NER. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (pp. 151–164). Association for Computational Linguistics.
+* Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V.. (2019). RoBERTa: A Robustly Optimized BERT Pretraining Approach. 
+