Whisper Small Finetuned Iwslt Irish Transfer Welsh

--- library_name: transformers license: apache-2.0 base_model: openai/whisper-small tags: - generated_from_trainer metrics: - wer - bleu model-index: - name: whisper-small-finetuned-iwslt-irish-transfer-welsh results: [] --- # whisper-small-finetuned-iwslt-irish-transfer-welsh This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the None dataset. It achieves the following results on the evaluation set: - Loss: 5.0553 - Wer: 106.7040 - Bleu: 2.6249 - Chrf++: 17.6130 - Comet: 0.4422 ## 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: 5e-05 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - optimizer: Use OptimizerNames.ADAMW_TORCH 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: 1000 - training_steps: 5000 - mixed_precision_training: Native AMP - label_smoothing_factor: 0.1 ### Training results | Training Loss | Epoch | Step | Validation Loss | Wer | Bleu | Chrf++ | Comet | |:-------------:|:-------:|:----:|:---------------:|:--------:|:------:|:-------:|:------:| | 2.2321 | 2.6709 | 1250 | 4.5102 | 107.1196 | 2.0090 | 16.9826 | 0.4377 | | 1.5919 | 5.3419 | 2500 | 4.7604 | 106.0625 | 2.4891 | 17.7932 | 0.4434 | | 1.5177 | 8.0128 | 3750 | 4.9138 | 106.1348 | 2.6324 | 17.7533 | 0.4397 | | 1.461 | 10.6838 | 5000 | 5.0553 | 106.7040 | 2.6249 | 17.6130 | 0.4422 | ### Framework versions - Transformers 4.52.0 - Pytorch 2.12.1+cu130 - Datasets 2.21.0 - Tokenizers 0.21.4