File size: 999 Bytes
a80fc49 6cb8e83 a80fc49 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 | ---
datasets:
- jstet/quotes-500k
language:
- en
metrics:
- perplexity
tags:
- text-generation-inference
---
# Quotes Generator
## Project description
Fine-tuned GPT2 model on the Quotes-500K dataset. For a given user prompt, it can generate motivational quotes starting with it. The model weights are deployed on Hugging Face Models Hub.
Check it out at https://huggingface.co/nandinib1999/quote-generator.
## Training data
This is the distribution of the total dataset into training, validation and test dataset for the fine-tuning task.
<table style="width:30%">
<tr>
<th>train</th>
<td>349796</td>
</tr>
<tr>
<th>validation</th>
<td>99942</td>
</tr>
<tr>
<th>test</th>
<td>49971</td>
</tr>
</table>
## Training procedure
The model was fine-tuned using the Google Colab GPU for one epoch. The weights of the pre-trained GPT2 model were used as a base.
## Eval results
<table style="width:30%">
<tr>
<th>Epoch</th>
<th>Perplexity</th>
</tr>
<tr>
<td>1</td>
<td>15.180</td>
</tr>
</table> |