whisper-ro
This model is a fine-tuned version of openai/whisper-small on the Echo dataset, a large open-source Romanian dataset.
Name | Small | Large-v2 | Fine-tuned small (this model) |
---|---|---|---|
Common Voice | 33.2 | 15.8 | 12.2 |
FLEURS | 29.8 | 14.4 | 10.9 |
VoxPopuli | 28.6 | 14.4 | 9.4 |
Echo | >100 | >100 | 8.6 |
RSC | 38.6 | 28.5 | 5.4 |
Training hyperparameters
The following hyperparameters were used during training:
learning_rate
: 1e-05train_batch_size
: 128eval_batch_size
: 128seed
: 42distributed_type
: multi-GPUnum_devices
: 2total_train_batch_size
: 256total_eval_batch_size
: 256optimizer
: Adam with betas=(0.9,0.999) and epsilon=1e-08lr_scheduler_type
: linearlr_scheduler_warmup_steps
: 500num_epochs
: 20.0mixed_precision_training
: Native AMP
- Downloads last month
- 4
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.