whisper-tiny-khmer-aug-kcc-v3

This model is a fine-tuned version of openai/whisper-tiny on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3416
• Wer: 100.0

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• 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: constant
• lr_scheduler_warmup_steps: 1000
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.1553	1.0	859	0.4808	100.0
0.4889	2.0	1718	0.3681	100.0
0.3862	3.0	2577	0.3386	100.0
0.3256	4.0	3436	0.3141	100.0
0.2824	5.0	4295	0.3053	100.0
0.2483	6.0	5154	0.3025	100.0
0.2191	7.0	6013	0.3178	100.0
0.1943	8.0	6872	0.3253	100.0
0.1747	9.0	7731	0.3295	100.0
0.1567	10.0	8590	0.3416	100.0

Framework versions

• Transformers 4.44.0
• Pytorch 2.3.1
• Datasets 2.21.0
• Tokenizers 0.19.1