flima/openai-whisper-large-v2-fullFT-es_ecu911_V2martin_win30s_samples

Whisper openai-whisper-large-v2

This model is a fine-tuned version of openai/whisper-large-v2 on the llamadas ecu911 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2832
• Wer: 19.8947

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: 2e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.5257	1.0	422	0.3596	35.3561
0.2108	2.0	844	0.2888	27.7362
0.1153	3.0	1266	0.2694	25.3256
0.0674	4.0	1688	0.2810	32.9731
0.0435	5.0	2110	0.2817	20.3380
0.0301	6.0	2532	0.2910	23.0673
0.0245	7.0	2954	0.2817	20.5043
0.021	8.0	3376	0.2734	18.4400
0.0173	9.0	3798	0.2784	23.9263
0.0167	10.0	4220	0.2832	19.8947

Framework versions

• Transformers 4.45.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.1