---
library_name: transformers
base_model: CausalNLP/gpt2-hf_multilingual-20
tags:
- generated_from_trainer
model-index:
- name: gpt2-multilingual-20-deu-repair_3epochs
  results: []
---

<!-- 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-multilingual-20-deu-repair_3epochs

This model is a fine-tuned version of [CausalNLP/gpt2-hf_multilingual-20](https://huggingface.co/CausalNLP/gpt2-hf_multilingual-20) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.2187

## 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: 32
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 500
- num_epochs: 3

### Training results

| Training Loss | Epoch  | Step  | Validation Loss |
|:-------------:|:------:|:-----:|:---------------:|
| 3.2849        | 0.0876 | 500   | 3.3025          |
| 3.2491        | 0.1752 | 1000  | 3.2839          |
| 3.2539        | 0.2628 | 1500  | 3.2729          |
| 3.2063        | 0.3504 | 2000  | 3.2656          |
| 3.2336        | 0.4380 | 2500  | 3.2591          |
| 3.2188        | 0.5256 | 3000  | 3.2539          |
| 3.2212        | 0.6132 | 3500  | 3.2500          |
| 3.2204        | 0.7008 | 4000  | 3.2457          |
| 3.2059        | 0.7884 | 4500  | 3.2427          |
| 3.2236        | 0.8760 | 5000  | 3.2396          |
| 3.2001        | 0.9636 | 5500  | 3.2368          |
| 3.1861        | 1.0512 | 6000  | 3.2354          |
| 3.1757        | 1.1388 | 6500  | 3.2339          |
| 3.1804        | 1.2264 | 7000  | 3.2319          |
| 3.1862        | 1.3140 | 7500  | 3.2297          |
| 3.1533        | 1.4016 | 8000  | 3.2279          |
| 3.1724        | 1.4892 | 8500  | 3.2262          |
| 3.1641        | 1.5768 | 9000  | 3.2249          |
| 3.1587        | 1.6644 | 9500  | 3.2235          |
| 3.1577        | 1.7520 | 10000 | 3.2221          |
| 3.1716        | 1.8396 | 10500 | 3.2212          |
| 3.2024        | 1.9272 | 11000 | 3.2203          |
| 3.1689        | 2.0147 | 11500 | 3.2199          |
| 3.1569        | 2.1023 | 12000 | 3.2198          |
| 3.1717        | 2.1899 | 12500 | 3.2195          |
| 3.1449        | 2.2775 | 13000 | 3.2192          |
| 3.1487        | 2.3651 | 13500 | 3.2190          |
| 3.1555        | 2.4527 | 14000 | 3.2189          |
| 3.1932        | 2.5403 | 14500 | 3.2188          |
| 3.1394        | 2.6279 | 15000 | 3.2187          |
| 3.1678        | 2.7155 | 15500 | 3.2187          |
| 3.1385        | 2.8032 | 16000 | 3.2187          |
| 3.1675        | 2.8908 | 16500 | 3.2187          |
| 3.183         | 2.9784 | 17000 | 3.2187          |


### Framework versions

- Transformers 4.57.3
- Pytorch 2.9.0
- Datasets 4.4.1
- Tokenizers 0.22.1