steja/whisper-small-occitan

Whisper_small_Occitan

This model is a fine-tuned version of bofenghuang/whisper-small-cv11-french on the google/fleurs oc_fr dataset. It achieves the following results on the evaluation set:

• Loss: 1.1744
• Wer: 39.8485

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 16
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• total_eval_batch_size: 64
• 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
0.0087	30.77	400	1.1744	39.8485
0.001	61.54	800	1.2807	39.8939
0.0005	92.31	1200	1.3227	40.3447
0.0004	123.08	1600	1.3445	40.1098
0.0003	153.85	2000	1.3524	40.0492

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu117
• Datasets 2.7.1
• Tokenizers 0.13.2