bert-base-sst2-finetuned

BERT Base Uncased – Fine-tuned on SST-2

This model is a fine-tuned version of bert-base-uncased on the GLUE SST-2 dataset for binary sentiment analysis.

---

Model Details

• Developed by: Talip7
• Base model: BERT Base Uncased
• Model type: Transformer encoder (BERT)
• Language: English
• Task: Sentiment Analysis (Binary Classification)

---

Training Details

• Dataset: GLUE / SST-2
• Training framework: PyTorch
• Libraries: 🤗 Transformers, 🤗 Datasets, 🤗 Accelerate
• Optimizer: AdamW
• Learning rate: 3e-5
• Epochs: 3
• Learning rate scheduler: Linear
• Hardware: GPU (via 🤗 Accelerate)

---

Evaluation Results

The model was evaluated on the SST-2 validation set.

• Accuracy: 0.9289 (92.89%)

---

Intended Use

This model can be used for:

• Binary sentiment analysis on English text
• Educational purposes (learning fine-tuning with Hugging Face)
• Benchmarking sentiment classification models

---

Limitations

• Trained only on movie reviews (SST-2); performance may degrade on other domains.
• Does not explicitly handle sarcasm or complex sentiment.
• Not suitable for multilingual sentiment analysis.

---

Usage

🤗 Transformers Pipeline

from transformers import pipeline

classifier = pipeline(
    "text-classification",
    model="Talip7/bert-base-sst2-finetuned"
)

classifier("I love this project!")

🔥 PyTorch Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

tokenizer = AutoTokenizer.from_pretrained("Talip7/bert-base-sst2-finetuned")
model = AutoModelForSequenceClassification.from_pretrained("Talip7/bert-base-sst2-finetuned")

text = "This movie was absolutely fantastic!"

inputs = tokenizer(text, return_tensors="pt")

with torch.no_grad():
    outputs = model(**inputs)

logits = outputs.logits
prediction = torch.argmax(logits, dim=-1).item()

label_map = {0: "Negative", 1: "Positive"}
print(f"Prediction: {label_map[prediction]}")