MCV_Fleurs_Combined_Irish_ASR_No_Aug2_No_lrscheduler

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

• Loss: 0.3382
• Wer: 29.6059

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 8
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.3585	0.9174	200	0.5024	54.8340
0.6899	1.8349	400	0.3746	32.6697
0.3285	2.7523	600	0.3440	29.9569
0.1760	3.6697	800	0.3269	29.8415
0.1083	4.5872	1000	0.3276	29.1144
0.0850	5.5046	1200	0.3319	28.1767
0.0673	6.4220	1400	0.3316	27.9811
0.0579	7.3394	1600	0.3311	27.6602
0.0662	8.0	1744	0.3382	29.6059

Framework versions

• Transformers 5.6.0.dev0
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.2