Hums003/distilbert-imdb-sentiment

DistilBERT Sentiment Analysis Model

This model is a fine-tuned version of distilbert-base-uncased for binary sentiment classification on the IMDB movie reviews dataset.

Model Details

Model Description

• Model type: DistilBERT (transformer-based)
• Task: Binary sentiment classification (positive/negative)
• Base Model: distilbert-base-uncased
• Language: English

Training Details

Training Data

• Dataset: IMDB Movie Reviews
• Training Samples: 16,000
• Validation Samples: 4,000
• Test Samples: 5,000
• Class Distribution: 50% positive, 50% negative

Training Procedure

• Epochs: 3
• Batch Size: 16
• Learning Rate: 2e-05
• Max Sequence Length: 512
• Optimizer: AdamW with weight decay (0.01)
• Scheduler: Linear with 10% warmup

Evaluation Results

• Test Accuracy: 0.9460
• Test F1 Score: 0.9723
• Best Validation Accuracy: 0.9300
• Training Time: ~6 minutes on Google Colab T4 GPU

How to Use

Direct Inference

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

# Load model and tokenizer
model_name = "Hums003/distilbert-imdb-sentiment"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)

# Prepare text
text = "This movie was absolutely fantastic! I loved every minute of it."
inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=512)

# Get predictions
with torch.no_grad():
    outputs = model(**inputs)
    predictions = torch.nn.functional.softmax(outputs.logits, dim=-1)

# Interpret results
sentiment = "positive" if predictions[0][1] > 0.5 else "negative"
confidence = predictions[0][1].item() if predictions[0][1] > 0.5 else predictions[0][0].item()
print(f"Sentiment: {sentiment} (confidence: {confidence:.2%})")