README.md

---
language: en
library_name: transformers
pipeline_tag: summarization
tags:
- youtube
- comments
- summarization
datasets:
- sujayC66/text_summarization_512_length_1_4000
metrics:
- rouge
base_model:
- Sivakkanth/youtube_comments_summarizer
---

# YouTube Comments Summarizer

This model is fine-tuned to summarize YouTube comments into a concise summary.  
It is based on **T5** and can be used directly with the Hugging Face `transformers` pipeline.

---

## Usage Example

```python
from transformers import pipeline

# Load the summarization pipeline from Hugging Face
model_id = "Sivakkanth/youtube_comments_summarizer"

summarizer = pipeline("summarization", model=model_id, tokenizer=model_id)

# Sample YouTube comment text
comments_text = """
This is a really interesting video about natural language processing. 
I learned a lot about different techniques for text summarization. 
The presenter explained everything clearly and the examples were helpful. 
I would recommend this video to anyone interested in NLP.
"""

# Generate summary
result = summarizer(
    comments_text,
    max_length=128,
    min_length=30,
    do_sample=False
)

print("Original Text:")
print(comments_text)
print("\nGenerated Summary:")
print(result[0]['summary_text'])


Input:
This is a really interesting video about natural language processing. 
I learned a lot about different techniques for text summarization. 
The presenter explained everything clearly and the examples were helpful. 
I would recommend this video to anyone interested in NLP.


Output (example):
This video about NLP was very informative and clearly explained, with helpful examples.

## Eval Results

Evaluation on a held-out YouTube comments test set:

- **ROUGE-1:** 0.5676652376831697  
- **ROUGE-2:** 0.3758989832045812
- **ROUGE-L:** 0.4824726190654699