lorenzoncina/whisper-large-ar

Whisper Large Arabic

This model is a fine-tuned version of openai/whisper-large on the mozilla-foundation/common_voice_11_0 ar dataset. It achieves the following results on the evaluation set:

• Loss: 0.3231
• Wer: 49.4320

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: 2
• 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: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2472	0.1	1000	0.3719	58.9560
0.2015	0.2	2000	0.3487	53.5213
0.1418	1.04	3000	0.3231	49.4320
0.0921	1.14	4000	0.3284	56.1107
0.0923	1.24	5000	0.3304	61.4227
0.0483	2.08	6000	0.3460	55.952
0.0391	2.18	7000	0.3538	51.1067
0.0228	3.02	8000	0.3493	51.82
0.0206	3.12	9000	0.3729	52.4000
0.018	3.22	10000	0.3676	51.296

Framework versions

• Transformers 4.28.0.dev0
• Pytorch 2.0.0+cu117
• Datasets 2.11.1.dev0
• Tokenizers 0.13.2