kpriyanshu256/whisper-large-v2-cy-500-32-1e-05

openai/whisper-large-v2-welsh

This model is a fine-tuned version of openai/whisper-large-v2 on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2947
• Wer: 18.0609

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
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.4438	0.2	100	0.4208	27.3594
0.3255	0.4	200	0.3633	23.6118
0.2856	0.6	300	0.3248	20.7023
0.1811	1.14	400	0.3011	18.5534
0.1404	1.34	500	0.2947	18.0609

Framework versions

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