finetune_v1

This model is a fine-tuned version of openai/whisper-large-v3 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.8789
• Wer: 115.1042

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: 2
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 300
• training_steps: 2400
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0	300.0	300	0.1414	50.0
0.0	600.0	600	0.3828	28.125
0.0	900.0	900	0.7280	97.9167
0.0	1200.0	1200	1.1172	126.5625
0.0	1500.0	1500	1.4219	125.5208
0.0	1800.0	1800	1.6904	119.7917
0.0	2100.0	2100	1.9209	115.1042
0.0	2400.0	2400	1.8789	115.1042

Framework versions

• Transformers 4.41.2
• Pytorch 2.2.0
• Datasets 2.20.0
• Tokenizers 0.19.1