---
license: mit
datasets:
- samsum
language:
- en
---

# Llama-2-7b Fine-Tuned Summarization Model

## Overview

The Llama-2-7b Fine-Tuned Summarization Model is a language model fine-tuned for the task of text summarization using QLora.
It has been fine-tuned on the samsum dataset, which contains a wide variety of coversation.

## Model Details

- Base Model: [meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf)
- Fine-Tuned on: [samsum dataset](https://huggingface.co/datasets/samsum)
- Language: English

## How to Use

You can use this model for text summarization tasks by utilizing the Hugging Face Transformers library. Here's a basic example in Python:

```python
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig

model_id =  "SalmanFaroz/Llama-2-7b-samsum"

bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_use_double_quant=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_compute_dtype=torch.bfloat16
)

model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=bnb_config, device_map="auto")

tokenizer = AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token = tokenizer.eos_token
tokenizer.padding_side = "right"

# Define the input prompt
prompt = """
Summarize the following conversation.

### Input:
Itachi: Kakashi, you must understand the gravity of the situation. The Akatsuki's plans are far more sinister than you can imagine.
Kakashi: Itachi, I need more than vague warnings. Tell me what you know.
Itachi: Very well. The Akatsuki seeks to capture Naruto for the power of the Nine-Tails sealed within him, but there's an even darker secret lurking within their goals.
Kakashi: Darker than that? What are they truly after?
Itachi: They're hunting the Tailed Beasts for a cataclysmic plan to reshape the world, and only we can stop them, together.

### Summary:
"""

inputs = tokenizer(prompt, return_tensors='pt')
output = tokenizer.decode(
    model.generate(
        inputs["input_ids"],
        max_new_tokens=100,
    )[0],
    skip_special_tokens=True
)

print("Output:",output)