CSY1109/drone_test1

Drone test En - Siang Yi

This model is a fine-tuned version of openai/whisper-small on the drone command test3 dataset. It achieves the following results on the evaluation set:

• Loss: 0.6268
• Wer: 8.3333

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: 3e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use 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_ratio: 0.1
• training_steps: 200
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.86	10.0	10	0.6997	25.0
0.0773	20.0	20	0.1492	16.6667
0.0062	30.0	30	0.5699	8.3333
0.0028	40.0	40	0.5715	8.3333
0.0006	50.0	50	0.5512	8.3333
0.0228	60.0	60	0.6065	8.3333
0.0	70.0	70	0.5899	8.3333
0.0001	80.0	80	0.6822	8.3333
0.0	90.0	90	0.6161	8.3333
0.0	100.0	100	0.6305	8.3333
0.0	110.0	110	0.6301	8.3333
0.0	120.0	120	0.6296	8.3333
0.0	130.0	130	0.6284	8.3333
0.0	140.0	140	0.6283	8.3333
0.0	150.0	150	0.6281	8.3333
0.0	160.0	160	0.6270	8.3333
0.0	170.0	170	0.6280	8.3333
0.0	180.0	180	0.6280	8.3333
0.0	190.0	190	0.6269	8.3333
0.0	200.0	200	0.6268	8.3333

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0