DianaJin/starstar

jinkrsmodel

This model is a fine-tuned version of openai/whisper-medium on the DianaJin/starstar dataset. It achieves the following results on the evaluation set:

• Loss: 0.8269
• Cer: 9.2920

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: 16
• 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: 50
• training_steps: 400
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
0.0011	50.0	100	0.6455	9.2920
0.0003	100.0	200	0.7593	9.2920
0.0003	150.0	300	0.8092	9.2920
0.0003	200.0	400	0.8269	9.2920

Framework versions

• Transformers 4.36.0.dev0
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0