---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- gsm8k
model-index:
- name: flan-t5-base-finetuned-gsm8k
  results: []
widget:
- text: "Please, answer the following question reasoning step-by-step:
Manu bought 4 apples and lost one in the market. How many apples does Manu have?"
---

<!-- 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. -->

# flan-t5-base-finetuned-gsm8k

This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on the gsm8k dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3652
- Rouge2 Precision: 0.3914
- Rouge2 Recall: 0.0816
- Rouge2 Fmeasure: 0.1308

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5

### Training results

| Training Loss | Epoch | Step | Validation Loss | Rouge2 Precision | Rouge2 Recall | Rouge2 Fmeasure |
|:-------------:|:-----:|:----:|:---------------:|:----------------:|:-------------:|:---------------:|
| 0.425         | 1.0   | 1869 | 0.3942          | 0.3707           | 0.0774        | 0.1238          |
| 0.3849        | 2.0   | 3738 | 0.3769          | 0.3809           | 0.0795        | 0.1272          |
| 0.3663        | 3.0   | 5607 | 0.3698          | 0.3808           | 0.0805        | 0.1285          |
| 0.3553        | 4.0   | 7476 | 0.3659          | 0.3863           | 0.0805        | 0.129           |
| 0.3421        | 5.0   | 9345 | 0.3652          | 0.3914           | 0.0816        | 0.1308          |


### Framework versions

- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.2