---
license: mit
tags:
- peft
- lora
- tinyllama
- dialogue summarization
- text2text-generation
datasets:
- dialogsum
metrics:
- rouge
---

# TinyLlama Dialogue Summarization Fine-Tuned Model

This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) specifically for dialogue summarization tasks. It was fine-tuned using Parameter-Efficient Fine-Tuning (PEFT) with LoRA on the [DialogSum](https://huggingface.co/datasets/dialogsum) dataset.

## Model Details

* **Model Name:** TinyLlama-1.1B-Chat-v1.0 (fine-tuned)
* **Architecture:** Causal Language Model
* **Fine-tuning Method:** Parameter-Efficient Fine-Tuning (PEFT) with LoRA
* **Dataset:** DialogSum
* **Quantization:** 4-bit quantization was used during training to reduce memory consumption.

## Intended Uses

This model is intended for generating concise and accurate summaries of dialogues. It can be used for various applications, including:

* Summarizing customer service conversations.
* Generating meeting summaries.
* Creating summaries of chat logs.

## How to Use

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
peft_model_id = "artisokka/tinyllama-dialogsum-finetuned"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
model = PeftModel.from_pretrained(model, peft_model_id)

input_text = "Dialogue: ... (your dialogue here) ..."
input_ids = tokenizer(input_text, return_tensors="pt").input_ids.to(model.device)

outputs = model.generate(input_ids, max_length=200)
summary = tokenizer.decode(outputs[0], skip_special_tokens=True)

print(summary)