metadata
license: apache-2.0
base_model: Talha/URDU-ASR
tags:
- generated_from_trainer
datasets:
- common_voice_13_0
metrics:
- wer
model-index:
- name: URDU-ASR
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: common_voice_13_0
type: common_voice_13_0
config: ur
split: test
args: ur
metrics:
- name: Wer
type: wer
value: 1.0023598591821734
URDU-ASR
This model is a fine-tuned version of Talha/URDU-ASR on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:
- Loss: 3.1901
- Wer: 1.0024
- Cer: 0.9455
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: 7.5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.85,0.99) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 19.3758 | 0.59 | 25 | 8.7836 | 1.0 | 0.9999 |
| 6.0744 | 1.17 | 50 | 4.7540 | 1.0 | 0.9999 |
| 4.446 | 1.76 | 75 | 4.0785 | 1.0 | 0.9999 |
| 3.7656 | 2.34 | 100 | 3.5164 | 1.0024 | 0.9457 |
| 3.4626 | 2.93 | 125 | 3.3191 | 1.0024 | 0.9454 |
| 3.2974 | 3.51 | 150 | 3.2566 | 1.0024 | 0.9449 |
| 3.2203 | 4.1 | 175 | 3.2009 | 1.0024 | 0.9456 |
| 3.1955 | 4.69 | 200 | 3.1901 | 1.0024 | 0.9455 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1