2026-02-24_22-58-22

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

• Loss: 1.0902

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
3.0503	0.1123	500	1.3913
1.1839	0.2247	1000	1.1939
1.0958	0.3370	1500	1.1537
1.0801	0.4493	2000	1.1328
1.0594	0.5617	2500	1.1232
1.0647	0.6740	3000	1.1167
1.0342	0.7863	3500	1.1094
1.0445	0.8987	4000	1.1032
1.0301	1.0110	4500	1.1015
1.0363	1.1233	5000	1.0999
1.0189	1.2357	5500	1.0972
1.0208	1.3480	6000	1.0947
1.0301	1.4603	6500	1.0939
1.0225	1.5727	7000	1.0929
1.0236	1.6850	7500	1.0928
1.0169	1.7973	8000	1.0912
1.0197	1.9097	8500	1.0909
1.0317	2.0220	9000	1.0902
1.0072	2.1344	9500	1.0902
1.0213	2.2467	10000	1.0902

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1