---
library_name: transformers
license: apache-2.0
base_model: google/umt5-small
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: t5-asr-CV16
  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. -->

# t5-asr-CV16

This model is a fine-tuned version of [google/umt5-small](https://huggingface.co/google/umt5-small) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6678
- Wer: 0.7639

## 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: 32
- seed: 42
- gradient_accumulation_steps: 128
- total_train_batch_size: 4096
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer    |
|:-------------:|:-------:|:----:|:---------------:|:------:|
| 1.8105        | 1.9694  | 48   | 0.7812          | 0.8528 |
| 1.6752        | 3.9694  | 96   | 0.7174          | 0.8285 |
| 1.6146        | 5.9694  | 144  | 0.7357          | 0.8215 |
| 1.3847        | 7.9694  | 192  | 0.6796          | 0.8172 |
| 1.2792        | 9.9694  | 240  | 0.6601          | 0.7841 |
| 1.2129        | 11.9694 | 288  | 0.6540          | 0.7764 |
| 1.279         | 13.9694 | 336  | 0.6792          | 0.7837 |
| 1.1706        | 15.9694 | 384  | 0.6695          | 0.7888 |
| 1.0348        | 17.9694 | 432  | 0.6931          | 0.7948 |
| 0.9335        | 19.9694 | 480  | 0.6678          | 0.7639 |


### Framework versions

- Transformers 4.48.3
- Pytorch 2.5.1+cu121
- Datasets 2.17.1
- Tokenizers 0.21.0