---
license: mit
base_model: openai-community/gpt2
tags:
- generated_from_trainer
model-index:
- name: GPT2-small-finetuned-Maximofn-short-jokes-dataset-casualLM
  results: []
datasets:
- Maximofn/short-jokes-dataset
language:
- en
widget:
  - text: <SJ> Why didn't the frog cross the road?
pipeline_tag: text-generation
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# GPT2-small-finetuned-Maximofn-short-jokes-dataset-casualLM

This model is a fine-tuned version of [openai-community/gpt2](https://huggingface.co/openai-community/gpt2) on [Maximofn/short-jokes-dataset](https://huggingface.co/datasets/Maximofn/short-jokes-dataset) dataset.
It achieves the following results on the evaluation set:

It is the result of the post [Fine tunning SML](https://maximofn.com/fine-tuning-sml/)

- Loss: 3.2013

## Model description

This model generated english jokes

## Intended uses & limitations

Text generation, english jokes

## Training and evaluation data

It is training on [Maximofn/short-jokes-dataset](https://huggingface.co/datasets/Maximofn/short-jokes-dataset) dataset.

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 28
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- lr_scheduler_warmup_steps: 100
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss |
|:-------------:|:-----:|:-----:|:---------------:|
| 3.3866        | 1.0   | 7447  | 3.2590          |
| 3.2599        | 2.0   | 14894 | 3.1997          |
| 3.2126        | 3.0   | 22341 | 3.1920          |


### Framework versions

- Transformers 4.42.4
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1