wav2vec2-base-random-stop-classification-3

This model is a fine-tuned version of on the audiofolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.3662
• Accuracy: 0.8753

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6931	0.99	18	0.6542	0.6178
0.6854	1.97	36	0.6173	0.6696
0.6617	2.96	54	0.5338	0.7343
0.6747	4.0	73	0.6521	0.6757
0.5626	4.99	91	0.4320	0.8072
0.5127	5.97	109	0.4987	0.7834
0.486	6.96	127	0.3753	0.8467
0.4393	8.0	146	0.4076	0.8290
0.4191	8.99	164	0.3877	0.8454
0.4287	9.97	182	0.3613	0.8549
0.4161	10.96	200	0.3714	0.8556
0.3938	12.0	219	0.3561	0.8569
0.3736	12.99	237	0.3914	0.8583
0.3571	13.97	255	0.3917	0.8535
0.3711	14.96	273	0.4288	0.8222
0.3303	16.0	292	0.3680	0.8638
0.3355	16.99	310	0.3724	0.8631
0.3523	17.97	328	0.3741	0.8644
0.3384	18.96	346	0.3726	0.8597
0.3063	20.0	365	0.3705	0.8658
0.2984	20.99	383	0.3866	0.8604
0.2841	21.97	401	0.3897	0.8590
0.3057	22.96	419	0.3662	0.8699
0.2831	24.0	438	0.3627	0.8760
0.2863	24.66	450	0.3662	0.8753

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.0
• Datasets 2.7.1
• Tokenizers 0.13.2