cardiffnlp
/
twitter-xlm-roberta-base-sentiment
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Compare commits

base: cardiffnlp:856553beaaaf559c029023df1e1af05a35723b77
Branches Tags
cardiffnlp:main
..
compare: cardiffnlp:d38ef7b8626347c12363d3058a8db28fe6dbcdb4
Branches Tags
cardiffnlp:main

10 Commits
856553beaa ... d38ef7b862

Author SHA1 Message Date
Asahi Ushio d38ef7b862 Update config.json 2022-11-28 11:30:47 +00:00
Jose Camacho Collados f3e34b6c30 Update README.md 2022-06-22 19:15:32 +00:00
Cardiff NLP 88a0007355 Add reference paper 2021-04-27 15:40:10 +00:00
Cardiff NLP 083e90c883 Update README.md 2021-04-26 16:28:59 +00:00
Cardiff NLP 54186b74b2 Update README.md 2021-04-19 17:08:05 +00:00
Cardiff NLP 968fc69b26 Update README.md 2021-04-17 06:52:13 +00:00
Cardiff NLP 34227e9348 Update README.md 2021-04-17 06:43:06 +00:00
Cardiff NLP 806cad138d Update README.md 2021-04-17 06:35:23 +00:00
Cardiff NLP b8a60461af Update README.md 2021-04-17 06:30:28 +00:00
Cardiff NLP d80dae4ef8 Update README.md 2021-04-17 06:29:57 +00:00
2 changed files with 15 additions and 14 deletions
Show Stats Expand all files Collapse all files

17
README.md
View File

@ -1,21 +1,23 @@
--- ---
language: multilingual language: multilingual
widget: widget:
- text: "T'estimo!" - text: "🤗"
- text: "T'estimo! ❤️"
- text: "I love you!" - text: "I love you!"
- text: "I hate you" - text: "I hate you 🤮"
- text: "Mahal kita!" - text: "Mahal kita!"
- text: "사랑해!" - text: "사랑해!"
- text: "난 너가 싫어" - text: "난 너가 싫어"
- text: "😍😍😍"
--- ---
# twitter-XLM-roBERTa-base for Sentiment Analysis # twitter-XLM-roBERTa-base for Sentiment Analysis
This is a XLM-roBERTa-base model trained on ~198M tweets and finetuned for sentiment analysis in This is a multilingual XLM-roBERTa-base model trained on ~198M tweets and finetuned for sentiment analysis. The sentiment fine-tuning was done on 8 languages (Ar, En, Fr, De, Hi, It, Sp, Pt) but it can be used for more languages (see paper for details).
- Paper: [XLM-T: A Multilingual Language Model Toolkit for Twitter](https://...). - Paper: [XLM-T: A Multilingual Language Model Toolkit for Twitter](https://arxiv.org/abs/2104.12250).
- Git Repo: [Tweeteval official repository](https://github.com/cardiffnlp/xlm-t). - Git Repo: [XLM-T official repository](https://github.com/cardiffnlp/xlm-t).
## Example Pipeline ## Example Pipeline
```python ```python
@ -24,9 +26,8 @@ model_path = "cardiffnlp/twitter-xlm-roberta-base-sentiment"
sentiment_task = pipeline("sentiment-analysis", model=model_path, tokenizer=model_path) sentiment_task = pipeline("sentiment-analysis", model=model_path, tokenizer=model_path)
sentiment_task("T'estimo!") sentiment_task("T'estimo!")
``` ```
Output:
``` ```
[{'label': 'LABEL_2', 'score': 0.6600581407546997}] [{'label': 'Positive', 'score': 0.6600581407546997}]
``` ```
## Full classification example ## Full classification example
@ -34,7 +35,7 @@ Output:
```python ```python
from transformers import AutoModelForSequenceClassification from transformers import AutoModelForSequenceClassification
from transformers import TFAutoModelForSequenceClassification from transformers import TFAutoModelForSequenceClassification
from transformers import AutoTokenizer from transformers import AutoTokenizer, AutoConfig
import numpy as np import numpy as np
from scipy.special import softmax from scipy.special import softmax

12
config.json
View File

@ -11,16 +11,16 @@
"hidden_dropout_prob": 0.1, "hidden_dropout_prob": 0.1,
"hidden_size": 768, "hidden_size": 768,
"id2label": { "id2label": {
"0": "Negative", "0": "negative",
"1": "Neutral", "1": "neutral",
"2": "Positive" "2": "positive"
}, },
"initializer_range": 0.02, "initializer_range": 0.02,
"intermediate_size": 3072, "intermediate_size": 3072,
"label2id": { "label2id": {
"Negative": 0, "negative": 0,
"Neutral": 1, "neutral": 1,
"Positive": 2 "positive": 2
}, },
"layer_norm_eps": 1e-05, "layer_norm_eps": 1e-05,
"max_position_embeddings": 514, "max_position_embeddings": 514,