| # Model Card for Thai Sentiment Classifier | |
| <!-- Add a brief description of your model here --> | |
| This model is a fine-tuned sentiment classifier for the Thai language, based on the xlm-roberta-base. | |
| <!-- Add more sections below as needed, e.g., Model Details, Uses, Bias/Risks/Limitations, Evaluation, etc. --> | |
| <!-- You can copy content from the markdown templates you used previously --> | |
| ## Model Details | |
| ### Model Description | |
| This model is a fine-tuned version of the multilingual language model `xlm-roberta-base` for the task of Thai sentiment classification. It was trained on the `sentiment_102` dataset to classify Thai text into one of four categories: positive, negative, neutral, or question. | |
| - **Developed by:** ZombitX64 | |
| - **Model type:** [More Information Needed] | |
| - **Language(s) (NLP):** th | |
| - **License:** apache-2.0 | |
| - **Finetuned from model:** xlm-roberta-base | |
| ## Uses | |
| ### Direct Use | |
| This model is intended for direct use in applications requiring sentiment analysis of Thai text. This includes: | |
| - Classifying customer reviews or feedback. | |
| - Analyzing social media posts. | |
| - Sorting survey responses by sentiment. | |
| - Identifying questions in text data. | |
| ### Out-of-Scope Use | |
| This model is not intended for: | |
| - Analyzing sentiment in languages other than Thai. | |
| - Analyzing nuanced emotions or complex linguistic structures beyond basic sentiment categories. | |
| - Use in applications where misclassification could lead to significant harm or impact. | |
| ## Bias, Risks, and Limitations | |
| The model's performance is limited by the quality and diversity of the dataset it was trained on. It may exhibit biases present in the training data. Performance on texts significantly different from the training data may be reduced. | |
| ## Evaluation | |
| Evaluation metrics from training: | |
| - Accuracy: 0.85 | |