xbilek25/whisper-small-train-basic_1000_v1.1

basic_train_basic_test 1000 similar params: per_device_train_batch_size=8, # bylo 16 a pod tim 1 gradient_accumulation_steps=2, warmup_steps=300, max_steps=3000

This model is a fine-tuned version of openai/whisper-small on the xbilek25/basic_train_set_en_last3cs_1000 dataset. It achieves the following results on the evaluation set:

• Loss: 0.1368
• Wer: 15.8003

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 300
• training_steps: 3000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0021	9.01	600	0.1419	16.5590
0.0002	19.0	1200	0.1322	15.3842
0.0001	28.02	1800	0.1351	15.2912
0.0001	38.01	2400	0.1362	15.8639
0.0001	47.02	3000	0.1368	15.8003

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.15.2