VladS159/Whisper_medium_ro_VladS_2000_steps_multi_gpu_26_02_2024

Whisper Medium Ro - Sarbu Vlad - multi gpu

This model is a fine-tuned version of openai/whisper-medium on the Common Voice 16.1 + Romanian speech synthesis dataset. It achieves the following results on the evaluation set:

• Loss: 0.0807
• Wer: 8.3506

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: 10
• eval_batch_size: 10
• seed: 42
• distributed_type: multi-GPU
• num_devices: 3
• total_train_batch_size: 30
• total_eval_batch_size: 30
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1753	0.22	250	0.1553	14.6629
0.1227	0.43	500	0.1203	11.8855
0.1286	0.65	750	0.1053	10.8999
0.1129	0.86	1000	0.0969	10.4557
0.051	1.08	1250	0.0882	9.1050
0.0589	1.3	1500	0.0851	9.0138
0.05	1.51	1750	0.0832	8.7795
0.0492	1.73	2000	0.0807	8.3506

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.0
• Datasets 2.17.0
• Tokenizers 0.15.1