sgangireddy/whisper-base-cv-highLR-cs

Whisper base Czech CV high LR

This model is a fine-tuned version of openai/whisper-base on the mozilla-foundation/common_voice_11_0 cs dataset. It achieves the following results on the evaluation set:

• Loss: 0.4828
• Wer: 26.9148

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: 64
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0606	4.01	1000	0.4679	34.8189
0.0153	8.02	2000	0.5124	32.4451
0.0018	12.03	3000	0.4891	29.0068
0.0004	16.04	4000	0.4804	27.1210
0.0002	21.01	5000	0.4828	26.9148

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.0+cu117
• Datasets 2.7.1.dev0
• Tokenizers 0.13.2