roma-whisper-small-levantine

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

Whisper Small - Levantine Arabic (Roma F1 Assistant)

This model is a fine-tuned version of openai/whisper-small on the NADI 2025 Subtask 2 (Jordan & Palestine) dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.5629
• eval_wer: 37.5784
• eval_runtime: 320.5385
• eval_samples_per_second: 1.997
• eval_steps_per_second: 0.25
• epoch: 6.9444
• step: 1250

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: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 2000
• mixed_precision_training: Native AMP

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.3
• Tokenizers 0.22.2