Pageee/DistilFT-Frisian-10md

DistilFT-Frisian-10m

This model is a fine-tuned version of distil-small.en on the mozilla-foundation/common_voice_6_fy_NL dataset. It achieves the following results on the evaluation set:

• Loss: 4.1946
• Wer: 89.3637

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• training_steps: 3000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1154	33.3333	500	3.7440	105.5534
0.0107	66.6667	1000	4.0814	88.1911
0.0	100.0	1500	4.1314	90.7432
0.0	133.3333	2000	4.1700	89.8663
0.0	166.6667	2500	4.1878	89.6061
0.0	200.0	3000	4.1946	89.3637

Framework versions

• Transformers 4.41.0.dev0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1