wavlm-base_3

This model is a fine-tuned version of microsoft/wavlm-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6534
• Accuracy: 0.8974

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 2
• seed: 0
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.2236	1.24	100	12.8495	0.4467
0.0514	2.48	200	16.3078	0.2677
0.0	3.72	300	17.5651	0.2597
0.3252	4.95	400	15.0382	0.1912
1.0577	6.19	500	0.6534	0.8974
0.6973	7.43	600	0.7352	0.1026
0.6939	8.67	700	0.6210	0.8974
0.6944	9.91	800	0.7129	0.1026

Framework versions

• Transformers 4.34.0.dev0
• Pytorch 2.0.0.post302
• Datasets 2.14.5
• Tokenizers 0.13.3