metadata
license: mit
base_model: gpt2
tags:
- generated_from_trainer
model-index:
- name: GPT2-arithmetic-3digits
results: []
datasets:
- Nbeau/additiondataset_3digits_10000examples
widget:
- text: |-
Input:
561+372
Target:
inference:
parameters:
add_special_tokens: true
do_sample: false
GPT2-arithmetic-3digits
This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0721
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.116 | 1.0 | 125 | 0.1085 |
| 0.0825 | 2.0 | 250 | 0.0801 |
| 0.0745 | 3.0 | 375 | 0.0740 |
| 0.0733 | 4.0 | 500 | 0.0724 |
| 0.7246 | 5.0 | 625 | 0.1426 |
| 0.0726 | 6.0 | 750 | 0.0721 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.3.0
- Datasets 2.19.1
- Tokenizers 0.19.1