| # Koelectra-five-sentiment-classification | |
| Koelectra on hugging face Transformers for Psychological Counseling | |
| - [full project link](https://github.com/jiminAn/Capstone_2022) | |
| ## how to use | |
| ``` | |
| from transformers import ElectraModel, ElectraTokenizer | |
| model = ElectraModel.from_pretrained("withU/Koelectra-five-sentiment-classification") | |
| tokenizer = ElectraTokenizer.from_pretrained("withU/Koelectra-five-sentiment-classification") | |
| categories = "withU/Koelectra-five-sentiment-classification" # 카테고리, index 파일 | |
| sentence = "나는 방금 밥을 먹었다." | |
| inputs = tokenizer.encode(sentence, return_tensors="pt") | |
| outputs = model(**inputs) | |
| softmax_logit = nn.Softmax(outputs).dim | |
| softmax_logit = softmax_logit[0].squeeze() | |
| max_index = torch.argmax(softmax_logit).item() | |
| prediction = max_index | |
| print(sentence, categories[prediction]) | |
| ``` | |
| ## dataset finetuned on | |
| - [wellness dataset](https://aihub.or.kr/opendata/keti-data/recognition-laguage/KETI-02-006) | |
| - [chatbot data](https://jeongukjae.github.io/tfds-korean/datasets/korean_chatbot_qa_data.html) | |
| - [korean-hate-speech](https://github.com/kocohub/korean-hate-speech) | |
| ## references | |
| - [WelllnessConversation-LanguageModel](https://github.com/nawnoes/WellnessConversation-LanguageModel) | |
| - [KoELECTRA](https://github.com/monologg/KoELECTRA) | |