---
pipeline_tag: conversational
language:
- en
library_name: transformers
---

This is Conversational chatbot built on top of DialoGPT-large with the inclusion of Harry Potter scripts, downloaded from [Kaggle here](https://www.kaggle.com/datasets/gulsahdemiryurek/harry-potter-dataset). 
The script is merged from 3 Harry Potter movies

Thanks to Lynn Zhang for her [tutorial here](https://www.freecodecamp.org/news/discord-ai-chatbot/) that inspired me to build this chatbot.

## How to run the model

Due to limitation in cloud computing from Hugging Face it might not be able to run the deployed model here, so I download the model to run on my local HPC system.
Here is the script that I used to enable the 4 line chat:


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

tokenizer = AutoTokenizer.from_pretrained("vuminhtue/DialoGPT-large-HarryPotter3")

model = AutoModelForCausalLM.from_pretrained("vuminhtue/DialoGPT-large-HarryPotter3")


# Let's chat for 4 lines
for step in range(4):
    # encode the new user input, add the eos_token and return a tensor in Pytorch
    new_user_input_ids = tokenizer.encode(input(">> User:") + tokenizer.eos_token, return_tensors='pt')
    # print(new_user_input_ids)

    # append the new user input tokens to the chat history
    bot_input_ids = torch.cat([chat_history_ids, new_user_input_ids], dim=-1) if step > 0 else new_user_input_ids

    # generated a response while limiting the total chat history to 1000 tokens, 
    chat_history_ids = model.generate(
        bot_input_ids, max_length=200,
        pad_token_id=tokenizer.eos_token_id,  
        no_repeat_ngram_size=3,       
        do_sample=True, 
        top_k=10, 
        top_p=0.5,
        temperature=0.5
    )
    
    # pretty print last ouput tokens from bot
    print("HarryPotter_Bot: {}".format(tokenizer.decode(chat_history_ids[:, bot_input_ids.shape[-1]:][0], skip_special_tokens=True)))
```