metadata
language:
- multilingual
license: apache-2.0
base_model: openai/whisper-medium
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- test000
metrics:
- wer
model-index:
- name: model trenovan na en_de_en simi setu, nastaveni jazyka en overeni3
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: odpovidajici nazvu modelu
type: test000
args: 'config: ende, split: train'
metrics:
- name: Wer
type: wer
value: 1195.9954233409612
model trenovan na en_de_en simi setu, nastaveni jazyka en overeni3
This model is a fine-tuned version of openai/whisper-medium on the odpovidajici nazvu modelu dataset. It achieves the following results on the evaluation set:
- Loss: 1.5807
- Wer: 1195.9954
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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 500
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.4162 | 3.06 | 125 | 1.4030 | 431.0069 |
| 0.0691 | 7.05 | 250 | 1.5075 | 1301.4874 |
| 0.0118 | 11.05 | 375 | 1.5467 | 1338.4439 |
| 0.0062 | 15.04 | 500 | 1.5807 | 1195.9954 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.6.0+cu124
- Datasets 3.4.1
- Tokenizers 0.15.2