MCV_Fleurs_Combined_Irish_ASR_normalised

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: 0.2974
• Wer: 20.3841

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: 8
• eval_batch_size: 8
• seed: 42
• 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: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 50000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.4442	1.1468	2000	0.5472	49.8797
0.2428	2.2936	4000	0.4101	36.6011
0.1367	3.4404	6000	0.3669	33.3066
0.0835	4.5872	8000	0.3278	28.3522
0.0616	5.7339	10000	0.3123	27.4596
0.0302	6.8807	12000	0.3026	24.5763
0.0070	8.0275	14000	0.3027	23.4931
0.0077	9.1743	16000	0.2871	22.1392
0.0032	10.3211	18000	0.2756	21.2015
0.0093	11.4679	20000	0.2705	20.3691
0.0480	12.6147	22000	0.3576	28.8286
0.0267	13.7615	24000	0.3431	27.2891
0.0342	14.9083	26000	0.3340	24.7217
0.0249	16.0550	28000	0.3439	25.7396
0.0148	17.2018	30000	0.3411	24.9524
0.0133	18.3486	32000	0.3326	23.8692
0.0207	19.4954	34000	0.3181	24.0548
0.0116	20.6422	36000	0.3162	22.3548
0.0072	21.7890	38000	0.3165	22.6607
0.0080	22.9358	40000	0.3259	22.4150
0.0033	24.0826	42000	0.3126	21.5675
0.0036	25.2294	44000	0.3091	21.6628
0.0044	26.3761	46000	0.3015	21.2316
0.0004	27.5229	48000	0.2968	20.4042
0.0038	28.6697	50000	0.2974	20.3841

Framework versions

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