metadata
language:
- te
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- openslr
metrics:
- wer
model-index:
- name: whisper-small-telugu-large-data
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: openslr
type: openslr
config: null
split: None
metrics:
- name: Wer
type: wer
value: 123.97713870997006
whisper-small-telugu-large-data
This model is a fine-tuned version of openai/whisper-small on the openslr and the google/fleurs datasets. It achieves the following results on the evaluation set:
- Loss: 1.6348
- Wer: 123.9771
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: 64
- 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
- training_steps: 50
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
2.2343 | 0.11 | 25 | 1.8167 | 116.8103 |
1.9575 | 0.23 | 50 | 1.6348 | 123.9771 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2