---
language:
- en
license: mit
base_model: microsoft/speecht5_tts
tags:
- .
- generated_from_trainer
datasets:
- speecht5_imda_nsc_p1_p3
model-index:
- name: Speech T5 TTS English
  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. -->

# Speech T5 TTS English

This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the IMDA National Speech Corpus dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5098

## 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: 1e-05
- train_batch_size: 16
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 20.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss |
|:-------------:|:-----:|:-----:|:---------------:|
| 0.9797        | 0.15  | 1000  | 0.7896          |
| 0.8628        | 0.3   | 2000  | 0.7427          |
| 0.8362        | 0.45  | 3000  | 0.7141          |
| 0.8327        | 0.6   | 4000  | 0.7071          |
| 0.8096        | 0.75  | 5000  | 0.7001          |
| 0.8085        | 0.89  | 6000  | 0.6922          |
| 0.8025        | 1.04  | 7000  | 0.6826          |
| 0.7962        | 1.19  | 8000  | 0.6762          |
| 0.7441        | 1.34  | 9000  | 0.6396          |
| 0.7332        | 1.49  | 10000 | 0.6278          |
| 0.7298        | 1.64  | 11000 | 0.6202          |
| 0.7208        | 1.79  | 12000 | 0.6103          |
| 0.7021        | 1.94  | 13000 | 0.6055          |
| 0.6951        | 2.09  | 14000 | 0.6022          |
| 0.7017        | 2.24  | 15000 | 0.6003          |
| 0.694         | 2.39  | 16000 | 0.5946          |
| 0.6757        | 2.54  | 17000 | 0.5842          |
| 0.6796        | 2.68  | 18000 | 0.5866          |
| 0.6724        | 2.83  | 19000 | 0.5829          |
| 0.6691        | 2.98  | 20000 | 0.5757          |
| 0.6617        | 3.13  | 21000 | 0.5698          |
| 0.6702        | 3.28  | 22000 | 0.5682          |
| 0.6623        | 3.43  | 23000 | 0.5658          |
| 0.6499        | 3.58  | 24000 | 0.5616          |
| 0.6545        | 3.73  | 25000 | 0.5573          |
| 0.6491        | 3.88  | 26000 | 0.5516          |
| 0.6398        | 4.03  | 27000 | 0.5509          |
| 0.6311        | 4.18  | 28000 | 0.5468          |
| 0.637         | 4.32  | 29000 | 0.5427          |
| 0.6271        | 4.47  | 30000 | 0.5385          |
| 0.6273        | 4.62  | 31000 | 0.5358          |
| 0.6132        | 4.77  | 32000 | 0.5321          |
| 0.6186        | 4.92  | 33000 | 0.5323          |
| 0.6157        | 5.07  | 34000 | 0.5320          |
| 0.6133        | 5.22  | 35000 | 0.5376          |
| 0.6088        | 5.37  | 36000 | 0.5228          |
| 0.6129        | 5.52  | 37000 | 0.5248          |
| 0.609         | 5.67  | 38000 | 0.5206          |
| 0.6042        | 5.82  | 39000 | 0.5197          |
| 0.6074        | 5.96  | 40000 | 0.5264          |
| 0.6029        | 6.11  | 41000 | 0.5267          |
| 0.6001        | 6.26  | 42000 | 0.5163          |
| 0.5999        | 6.41  | 43000 | 0.5177          |
| 0.5994        | 6.56  | 44000 | 0.5150          |
| 0.5982        | 6.71  | 45000 | 0.5141          |
| 0.605         | 6.86  | 46000 | 0.5146          |
| 0.5939        | 7.01  | 47000 | 0.5144          |
| 0.5934        | 7.16  | 48000 | 0.5129          |
| 0.5982        | 7.31  | 49000 | 0.5128          |
| 0.5925        | 7.46  | 50000 | 0.5088          |
| 0.5933        | 7.61  | 51000 | 0.5117          |
| 0.5861        | 7.75  | 52000 | 0.5098          |
| 0.5931        | 7.9   | 53000 | 0.5121          |


### Framework versions

- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2