luigisaetta/whisper-atco2-largev2-v2

openai/whisper-large-v2

This model is a fine-tuned version of openai/whisper-large-v2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7993
• Wer: 21.2788

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: 8
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• total_eval_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 400
• training_steps: 800

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.007	8.33	100	0.5728	21.4885
0.0007	16.67	200	0.7017	22.1174
0.0003	25.0	300	0.7358	21.5933
0.0002	33.33	400	0.7598	21.5933
0.0002	41.67	500	0.7793	22.0126
0.0001	50.0	600	0.7896	22.0126
0.0001	58.33	700	0.7969	21.2788
0.0001	66.67	800	0.7993	21.2788

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.1
• Datasets 2.8.1.dev0
• Tokenizers 0.13.2