whisper-large-v3-cv-capes-fs024-IEEE-pt

This model is a fine-tuned version of openai/whisper-large-v3 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1045

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-06
• train_batch_size: 128
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 256
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 1000

Training results

Training Loss	Epoch	Step	Validation Loss
0.1585	0.2315	50	0.1324
0.0913	0.4630	100	0.1042
0.0758	0.6944	150	0.0975
0.0732	0.9259	200	0.0940
0.0522	1.1574	250	0.0931
0.0536	1.3889	300	0.0919
0.0505	1.6204	350	0.0915
0.0491	1.8519	400	0.0913
0.0426	2.0833	450	0.0930
0.0358	2.3148	500	0.0933
0.036	2.5463	550	0.0934
0.0347	2.7778	600	0.0934
0.0344	3.0093	650	0.0935
0.0271	3.2407	700	0.0976
0.0269	3.4722	750	0.0988
0.027	3.7037	800	0.1005
0.0278	3.9352	850	0.0992
0.0228	4.1667	900	0.1037
0.0214	4.3981	950	0.1042
0.0225	4.6296	1000	0.1045

Framework versions

• Transformers 4.50.2
• Pytorch 2.5.1+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2