finetune_v12

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: 0.5752
• Wer: 50.1283

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: 5e-06
• train_batch_size: 16
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 8
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10
• training_steps: 160
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	20.0	20	0.6631	28.0582
No log	40.0	40	0.5728	21.8991
0.1708	60.0	60	0.5718	24.1232
0.1708	80.0	80	0.5820	24.8931
0.0072	100.0	100	0.5815	39.7776
0.0072	120.0	120	0.5669	38.8366
0.0072	140.0	140	0.5718	56.4585
0.0002	160.0	160	0.5752	50.1283

Framework versions

• Transformers 4.42.3
• Pytorch 2.2.0
• Datasets 2.20.0
• Tokenizers 0.19.1