whisper-small-ru-v13t

This model is a fine-tuned version of openai/whisper-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2455
• Wer: 17.8205

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: 32
• eval_batch_size: 32
• 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: 100
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2313	0.5814	100	0.2559	19.2122
0.1418	1.1628	200	0.2271	18.1979
0.1324	1.7442	300	0.2223	17.4903
0.0797	2.3256	400	0.2223	17.4195
0.0788	2.9070	500	0.2205	17.4313
0.0495	3.4884	600	0.2301	17.5139
0.0439	4.0698	700	0.2316	17.5257
0.0337	4.6512	800	0.2417	17.9031
0.0257	5.2326	900	0.2435	17.9856
0.0242	5.8140	1000	0.2455	17.8205

Framework versions

• Transformers 4.49.0
• Pytorch 2.1.0+cu118
• Datasets 3.3.2
• Tokenizers 0.21.0