indic-whisper-vulnerable1

This model is a fine-tuned version of HariprasathSB/indic-whisper-vulnerable on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.0378
• Wer: 72.3914

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: 4
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0175	0.8811	200	0.9461	75.1567
0.0294	1.7621	400	0.9488	75.5710
0.0227	2.6432	600	0.9532	73.7349
0.0123	3.5242	800	0.9391	74.6865
0.0059	4.4053	1000	1.0054	73.3318
0.0032	5.2863	1200	1.0145	75.1008
0.0022	6.1674	1400	1.0115	73.1975
0.0011	7.0485	1600	1.0310	72.8168
0.0008	7.9295	1800	1.0431	72.4250
0.0002	8.8106	2000	1.0378	72.3914

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1