Added BERT v/s Raw Performance

[Soubeer]

Related Issue

This pull request tackles the need for a detailed comparison of different text classification methods on the IMDB dataset. By comparing raw text processing, pre-trained BERT, and fine-tuned BERT models, I aim to highlight their performance differences. This analysis will help us understand which approach works best for sentiment analysis and guide future enhancements. It provides valuable insights into the strengths and weaknesses of each method, helping to make informed decisions for upcoming projects in the repository.

About Issue #188

Description

In this pull request, I've conducted a comparative study of text classification methods on the IMDB dataset: raw text processing, pre-trained BERT, and fine-tuned BERT models. BERT, known for its deep bidirectional representations, is evaluated both in its pre-trained state and after fine-tuning on the IMDB dataset. Fine-tuning involves further training BERT on our specific dataset, enhancing its performance by capturing domain-specific nuances. The analysis includes training, validation, and performance metrics for each method, highlighting the strengths and weaknesses of each approach. This comparison aims to guide future development and optimization efforts in our repository, ensuring the use of the most effective text classification methods.

Type of PR

• Bug fix
• [✓ ] Feature enhancement
• Documentation update
• Other (specify): _______________

Screenshots / videos (if applicable)

[Attach any relevant screenshots or videos demonstrating the changes]

(Not really required,already present in notebook file)

Checklist:

• I have performed a self-review of my code
• I have read and followed the Contribution Guidelines.
• I have tested the changes thoroughly before submitting this pull request.
• I have provided relevant issue numbers, screenshots, and videos after making the changes.
• I have commented my code, particularly in hard-to-understand areas.

Additional context:

I have not added a READme and have made the whole thing inside a new folder named raw_bert_finetune..should I change it or add something?

[sanjay-kv]

this never helps , unlike machine learning repo we cant do contribution in this way. you have to integrate to the existing files or enhancement.

[github-actions]

This PR has been automatically closed due to inactivity from the owner for 15 days.