README.md

metadata

license: apache-2.0
base_model: facebook/wav2vec2-large-xlsr-53
tags:
  - automatic-speech-recognition
  - DewiBrynJones/banc-trawsgrifiadau-bangor-clean-with-ccv
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: wav2vec2-xlsr-53-ft-btb-ccv-cy
    results: []

wav2vec2-xlsr-53-ft-btb-ccv-cy

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the DEWIBRYNJONES/BANC-TRAWSGRIFIADAU-BANGOR-CLEAN-WITH-CCV - DEFAULT dataset. It achieves the following results on the evaluation set:

• Loss: 1.5563
• Wer: 0.8702

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: 0.0003
• train_batch_size: 128
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0.3096	100	3.5623	1.0
No log	0.6192	200	3.1175	1.0
No log	0.9288	300	3.2933	1.0
No log	1.2384	400	0.9765	0.6866
3.7398	1.5480	500	0.6787	0.4922
3.7398	1.8576	600	0.6015	0.4347
3.7398	2.1672	700	0.5699	0.4273
3.7398	2.4768	800	0.5419	0.3958
3.7398	2.7864	900	0.4971	0.3730
0.5228	3.0960	1000	0.4960	0.3543
0.5228	3.4056	1100	0.6977	0.4655
0.5228	3.7152	1200	0.7999	0.5316
0.5228	4.0248	1300	1.0526	0.6215
0.5228	4.3344	1400	1.1285	0.7572
0.9047	4.6440	1500	1.1593	0.7047
0.9047	4.9536	1600	2.0401	0.9667
0.9047	5.2632	1700	1.6264	0.8680
0.9047	5.5728	1800	1.5916	0.8627
0.9047	5.8824	1900	1.5764	0.8720
1.6715	6.1920	2000	1.5563	0.8702

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1