README.md

metadata

language:
  - en
widget:
  - text: >-
      Generate a dialogue between two people about the following topic: A local
      street market bustles with activity, #Person1# tries exotic food for the
      first time, and #Person2#, familiar with the cuisine, offers insights and
      recommendations. Dialogue:
    example_title: Street Market
  - text: >-
      Generate a dialogue between two people about the following topic: At a
      quiet park, #Person1# stumbles upon an eerie crime scene, and #Person2#, a
      detective, arrives and begins to unravel the mysterious circumstances of
      the murder. Dialogue:
    example_title: Crime Scene
tags:
  - dialogue
  - conversation-generator
  - flan-t5-base
  - fine-tuned
license: apache-2.0
datasets:
  - kaggle-dialogue-dataset

Omaratef3221/flan-t5-base-dialogue-generator

Model Description

This model is a fine-tuned version of Google's t5 specifically tailored for generating realistic and engaging dialogues or conversations. It has been trained to capture the nuances of human conversation, making it highly effective for applications requiring conversational AI capabilities.

GitHub Repo: https://github.com/omaratef3221/conversation_generator_LLM

Intended Use

Omaratef3221/flan-t5-base-dialogue-generator is ideal for developing chatbots, virtual assistants, and other applications where generating human-like dialogue is crucial. It can also be used for research and development in natural language understanding and generation.

How to Use

You can use this model directly with the transformers library as follows:

Download the model

model_name = "Omaratef3221/flan-t5-base-dialogue-generator"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

Use with example

from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

model_name = "Omaratef3221/flan-t5-base-dialogue-generator"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

prompt = '''
Generate a dialogue between two people about the following topic:
A local street market bustles with activity, #Person1# tries exotic food for the first time, and #Person2#, familiar with the cuisine, offers insights and recommendations.
Dialogue:
'''
# Generate a response to an input statement
input_ids = tokenizer(prompt, return_tensors='pt').input_ids
output = model.generate(input_ids, top_p = 0.6, do_sample=True, temperature = 1.2, max_length = 512)
print(tokenizer.decode(output[0], skip_special_tokens=True).replace('#Person2#:', '\n#Person2#:').replace('#Person1#:', '\n#Person1#:'))