SentimentBERT

Model Card for SentimentBERT

This model is a fine-tuned version of bert-base-uncased for sentiment analysis. It has been trained on the Sentiment140 Kaggle dataset, enabling it to classify text as positive or negative.

Model Details

Model Description

This model is fine-tuned using the bert-base-uncased architecture to perform sentiment analysis. It accepts text input and predicts whether the sentiment expressed in the text is positive or negative.

• Developed by: Debopam(Pritam) Dey
• Funded by [optional]: Not specified
• Shared by [optional]: Debopam(Pritam) Dey
• Model type: Sequence classification (binary sentiment analysis)
• Language(s) (NLP): English
• License: Apache 2.0
• Finetuned from model [optional]: bert-base-uncased

Model Sources [optional]

• Repository: SentimentBERT
• Demo [optional]: Coming Soon

Uses

Here’s how to use the model for sentiment analysis:

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

# Load the model and tokenizer from the Hugging Face model hub
mymodel = AutoModelForSequenceClassification.from_pretrained("pritam2014/SentimentBERT")
mytokenizer = AutoTokenizer.from_pretrained("pritam2014/SentimentBERT")

# Preprocess the text input
def preprocess_text(text):
    inputs = mytokenizer.encode_plus(
        text,
        max_length=50,
        padding='max_length',
        truncation=True,
        return_attention_mask=True,
        return_tensors='pt'
    )
    return inputs

# Predict sentiment
def make_prediction(text):
    inputs = preprocess_text(text)
    with torch.no_grad():
        outputs = mymodel(inputs['input_ids'], attention_mask=inputs['attention_mask'])
        logits = outputs.logits
        predicted_class_id = torch.argmax(logits).item()
    sentiment_labels = {0: 'Negative', 1: 'Positive'}
    return sentiment_labels[predicted_class_id]

# Example
text = "I love this product!"
print(make_prediction(text))  # Output: Positive

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

Direct Use

The model can be used for text classification tasks without additional fine-tuning.

from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("pritam2014/SentimentBERT")
model = AutoModelForSequenceClassification.from_pretrained("pritam2014/SentimentBERT")

from transformers import pipeline

# Initialize pipeline
sentiment_pipeline = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)

# Example input
tweets = [
    "I love this product!", 
    "I'm not happy with the service.", 
    "It's okay, could be better."
]

# Predict sentiment
results = sentiment_pipeline(tweets)
for tweet, result in zip(tweets, results):
    print(f"Tweet: {tweet}\nSentiment: {result['label']}, Score: {result['score']:.4f}\n")

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

Users can fine-tune the model on other sentiment datasets or adapt it for related tasks like emotion detection.

Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

The model is not suitable for multilingual sentiment analysis or highly nuanced text where sentiment depends on complex context.

Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

• The model may inherit biases present in the Sentiment140 dataset.
• It is designed for English text and may perform poorly on non-English or mixed-language text.

Recommendations

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

Use the model in scenarios where binary sentiment classification is sufficient. Avoid deploying it in critical systems without further testing for biases and limitations.

How to Get Started with the Model

Refer to the "Uses" section above to see the sample usage code. For more details, visit the Hugging Face Hub page.

Training Details

Training Data

<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->

The model was fine-tuned on the Sentiment140 dataset, which contains 1.6 million tweets labelled as positive or negative.

Training Procedure

<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->

• Optimizer: AdamW
• Batch size: 760
• Learning rate: 1e-5
• Epochs: 2
• Hardware: Kaggle T4 GPU

Preprocessing [optional]

[More Information Needed]

Training Hyperparameters

• Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->

Speeds, Sizes, Times [optional]

<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->

[More Information Needed]

Evaluation

<!-- This section describes the evaluation protocols and provides the results. -->

The model was evaluated on a validation split of the Sentiment140 dataset.

Testing Data, Factors & Metrics

Testing Data

<!-- This should link to a Dataset Card if possible. -->

[More Information Needed]

Factors

<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->

[More Information Needed]

Metrics

<!-- These are the evaluation metrics being used, ideally with a description of why. -->

[More Information Needed]

Results

[More Information Needed]

Summary

Model Examination [optional]

<!-- Relevant interpretability work for the model goes here -->

[More Information Needed]

Environmental Impact

<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

• Hardware Type: [More Information Needed]
• Hours used: [More Information Needed]
• Cloud Provider: Kaggle T4 GPU
• Compute Region: [More Information Needed]
• Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->

BibTeX:

@misc{pritam2014SentimentBERT, author = {Debopam(Pritam) Dey}, title = {SentimentBERT}, year = {2025}, publisher = {Hugging Face}, howpublished = {\url{https://huggingface.co/pritam2014/SentimentBERT}}, }

APA:

[More Information Needed]

Glossary [optional]

<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->

[More Information Needed]

More Information [optional]

The model performs well on short texts like tweets but may require further fine-tuning for longer or domain-specific text.

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

For questions or feedback, feel free to contact me via the Hugging Face repository or email at (letsdecode2014@gmail.com)