CRCD DistilBERT Sentiment Classifier
A fine-tuned DistilBERT model for sentiment analysis of English customer and product reviews.
Task
The model classifies reviews into three sentiment classes:
| Label ID | Sentiment |
|---|---|
| 0 | negative |
| 1 | neutral |
| 2 | positive |
Model architecture
- Base model: distilbert/distilbert-base-uncased
- Model type: DistilBertForSequenceClassification
- Encoder: DistilBERT
- Number of output classes: 3
- Task: English sentiment classification
- Maximum input length: 256 tokens
The model contains a DistilBERT encoder and a classification head with three output neurons.
Training dataset
The model was fine-tuned on the CRCD Balanced Sentiment Dataset:
https://huggingface.co/datasets/SergeiM89/CRCD-sentiment-balanced-3class
The dataset contains English customer and product reviews with negative, neutral, and positive sentiment labels.
The data was cleaned, balanced, and divided into training, validation, and test splits before fine-tuning.
Related resources
Base model:
https://huggingface.co/distilbert/distilbert-base-uncased
Training dataset:
https://huggingface.co/datasets/SergeiM89/CRCD-sentiment-balanced-3class
- Downloads last month
- 24
Model tree for SergeiM89/CRCD-distilbert-sentiment-trained
Base model
distilbert/distilbert-base-uncased