Breeze DSW Telugu - base

This model is a fine-tuned version of openai/whisper-base on the google/fleurs te_in dataset. It achieves the following results on the evaluation set:

Loss: 0.3372
Wer: 37.4544

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: 32
eval_batch_size: 16
seed: 42
distributed_type: multi-GPU
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
training_steps: 2000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2937	2.03	200	0.3237	42.5614
0.1611	5.02	400	0.2756	38.9148
0.0889	8.01	600	0.2930	38.1106
0.0456	11.0	800	0.3372	37.4544
0.0229	13.03	1000	0.3982	37.9258
0.0103	16.02	1200	0.4473	38.2678
0.0042	19.02	1400	0.4836	37.8980
0.0025	22.01	1600	0.5083	37.7317
0.002	24.04	1800	0.5220	37.8010
0.0018	27.03	2000	0.5269	37.9027