metadata
language:
- ur
license: apache-2.0
tags:
- automatic-speech-recognition
- robust-speech-event
datasets:
- common_voice
metrics:
- wer
model-index:
- name: wav2vec2-large-xlsr-53-urdu
results:
- task:
type: automatic-speech-recognition
name: Urdu Speech Recognition
dataset:
type: common_voice
name: Urdu
args: ur
metrics:
- type: wer
value: 100
name: Test WER
args:
- learning_rate: 0.0003
- train_batch_size: 16
- eval_batch_size: 8
- 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: 10
- num_epochs: 30
- mixed_precision_training: Native AMP
wav2vec2-large-xlsr-53-urdu
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice dataset. It achieves the following results on the evaluation set:
- Loss: 2.6772
- Wer: 1.0
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: 0.0003
- train_batch_size: 16
- eval_batch_size: 8
- 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: 10
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 11.1125 | 3.33 | 40 | 3.2875 | 1.0 |
| 3.2077 | 6.67 | 80 | 3.1499 | 1.0 |
| 3.1725 | 10.0 | 120 | 3.1484 | 1.0 |
| 3.148 | 13.33 | 160 | 3.0948 | 1.0 |
| 3.1098 | 16.67 | 200 | 3.0897 | 1.0 |
| 3.085 | 20.0 | 240 | 3.0609 | 1.0 |
| 3.0315 | 23.33 | 280 | 2.9636 | 1.0 |
| 2.9038 | 26.67 | 320 | 2.7838 | 1.0 |
| 2.7599 | 30.0 | 360 | 2.6772 | 1.0 |
Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.17.0
- Tokenizers 0.10.3