whisper-large-v3-Kannada-Version1

This model is a fine-tuned version of openai/whisper-large-v3 on the fleurs dataset. It achieves the following results on the evaluation set:

• Loss: 0.1214
• Wer: 41.3722

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: 3e-06
• train_batch_size: 8
• 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: 1000
• training_steps: 20000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1966	6.0606	2000	0.1678	54.0036
0.1699	12.1212	4000	0.1455	48.0278
0.1607	18.1818	6000	0.1358	45.7829
0.1497	24.2424	8000	0.1304	43.5934
0.1413	30.3030	10000	0.1270	42.7713
0.146	36.3636	12000	0.1248	41.9730
0.1309	42.4242	14000	0.1233	41.6726
0.1339	48.4848	16000	0.1222	41.4987
0.1343	54.5455	18000	0.1218	41.5382
0.1267	60.6061	20000	0.1214	41.3722

Framework versions

• PEFT 0.12.1.dev0
• Transformers 4.45.0.dev0
• Pytorch 2.4.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1