---
library_name: transformers
language:
- en
license: mit
base_model: microsoft/speecht5_tts
tags:
- generated_from_trainer
datasets:
- lj_speech
model-index:
- name: SpeechT5 using custom dataset
  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. -->

# SpeechT5 using custom dataset

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

## 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: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- 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_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch     | Step | Validation Loss |
|:-------------:|:---------:|:----:|:---------------:|
| 1.7065        | 666.6667  | 1000 | nan             |
| 1.4393        | 1333.3333 | 2000 | nan             |
| 1.2369        | 2000.0    | 3000 | nan             |
| 1.1759        | 2666.6667 | 4000 | nan             |


### Framework versions

- Transformers 4.47.0.dev0
- Pytorch 2.5.0+cu121
- Datasets 3.0.2
- Tokenizers 0.20.1