GPT-2 for Dialogue

This repository contains gpt2 (the standard 124M parameter version) fine-tuned for improved conversational performance. The model was trained on a diverse, custom dataset of English dialogues.

Original model https://huggingface.co/openai-community/gpt2

Project Goal: The primary objective is to push the capabilities of the small GPT-2 architecture for conversational tasks, including general chat and question answering. The focus is on creating a capable yet lightweight dialogue model. This model represents an initial stage, with further training planned.

Training Data: A custom corpus of English dialogues spanning multiple general topics was used for fine-tuning.

Potential Use Cases:

Casual conversation partner
Base for further fine-tuning on specific dialogue tasks
Research into efficient conversational models

Disclaimer: This model is a work in progress. As it's based on gpt2, it may not perform as well as larger models and may exhibit limitations common to language models of this size.

Downloads last month: 16

GGUF

Model size

0.1B params

Architecture

gpt2

Hardware compatibility

8-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support