Configuration Parsing Warning:In adapter_config.json: "peft.task_type" must be a string

whisper-large-v3-turbo-med-pl-lora-r64-enc-dec-lr1e-04-ep7-whisper_bigos_all_fair

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

  • Loss: 1.6231
  • Model Preparation Time: 0.0193
  • Wer: 13.3673
  • Cer: 4.5623

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: 0.00015
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 7
  • mixed_precision_training: Native AMP
  • label_smoothing_factor: 0.1

Training results

Training Loss Epoch Step Validation Loss Model Preparation Time Wer Cer
1.6564 1.0 1827 1.6525 0.0193 14.2827 4.8698
1.625 2.0 3654 1.6305 0.0193 15.5315 6.6584
1.5918 3.0 5481 1.6233 0.0193 13.4289 4.6757
1.5743 4.0 7308 1.6240 0.0193 14.3337 5.4391
1.5544 5.0 9135 1.6220 0.0193 12.9638 4.4558
1.5415 6.0 10962 1.6210 0.0193 12.8512 4.1801
1.5333 7.0 12789 1.6231 0.0193 13.3673 4.5623

Framework versions

  • PEFT 0.18.1
  • Transformers 4.57.6
  • Pytorch 2.8.0+cu128
  • Datasets 4.5.0
  • Tokenizers 0.22.2
Downloads last month
85
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for AleksanderObuchowski/whisper-large-v3-turbo-med-pl-lora-r64-enc-dec-lr1e-04-ep7-whisper_bigos_all_fair

Adapter
(120)
this model