metadata
tags:
- generated_from_trainer
datasets:
- common_voice_17_0
metrics:
- wer
model-index:
- name: w2v2-bert-urdu
results:
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: common_voice_17_0
type: common_voice_17_0
config: ur
split: test[:100]
args: ur
metrics:
- type: wer
value: 0.3300546448087432
name: Wer
w2v2-bert-urdu
This model was trained from scratch on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4246
- Wer: 0.3301
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: 5e-06
- 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: 100
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.8145 | 0.1695 | 50 | 0.4620 | 0.3421 |
| 0.4364 | 0.3390 | 100 | 0.3969 | 0.2874 |
| 0.418 | 0.5085 | 150 | 0.3697 | 0.2820 |
| 0.402 | 0.6780 | 200 | 0.3627 | 0.2842 |
| 0.3698 | 0.8475 | 250 | 0.3314 | 0.2710 |
| 0.3779 | 1.0169 | 300 | 0.3292 | 0.2852 |
| 0.3167 | 1.1864 | 350 | 0.3230 | 0.2820 |
| 0.3578 | 1.3559 | 400 | 0.3825 | 0.2940 |
| 0.4189 | 1.5254 | 450 | 0.4225 | 0.3104 |
| 0.4803 | 1.6949 | 500 | 0.4248 | 0.3311 |
| 0.4612 | 1.8644 | 550 | 0.4246 | 0.3301 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1