weepower/CustomASRModel

1---2library_name: transformers3license: apache-2.04base_model: openai/whisper-small5tags:6- generated_from_trainer7metrics:8- wer9model-index:10- name: CustomASRModel11  results: []12---13 14<!-- This model card has been generated automatically according to the information the Trainer had access to. You15should probably proofread and complete it, then remove this comment. -->16 17# CustomASRModel18 19This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on an unknown dataset.20It achieves the following results on the evaluation set:21- Loss: 1.459322- Wer Ortho: 70.607923- Wer: 55.430924 25## Model description26 27More information needed28 29## Intended uses & limitations30 31More information needed32 33## Training and evaluation data34 35More information needed36 37## Training procedure38 39### Training hyperparameters40 41The following hyperparameters were used during training:42- learning_rate: 1e-0543- train_batch_size: 444- eval_batch_size: 1645- seed: 4246- gradient_accumulation_steps: 447- total_train_batch_size: 1648- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments49- lr_scheduler_type: constant_with_warmup50- lr_scheduler_warmup_steps: 5051- training_steps: 50052- mixed_precision_training: Native AMP53 54### Training results55 56| Training Loss | Epoch  | Step | Validation Loss | Wer Ortho | Wer     |57|:-------------:|:------:|:----:|:---------------:|:---------:|:-------:|58| 3.0148        | 0.8309 | 500  | 1.4593          | 70.6079   | 55.4309 |59 60 61### Framework versions62 63- Transformers 5.2.064- Pytorch 2.10.0+cu13065- Datasets 4.6.066- Tokenizers 0.22.267