wav2vec2-base-random-stop-classification-2

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

• Loss: 0.4265
• Accuracy: 0.8569

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
0.6925	0.99	18	0.6506	0.6049
0.6667	1.97	36	0.6474	0.6396
0.5762	2.96	54	0.5791	0.7670
0.559	4.0	73	0.4603	0.7963
0.4892	4.99	91	0.4248	0.8161
0.4853	5.97	109	0.4544	0.8113
0.4452	6.96	127	0.5181	0.8011
0.4747	8.0	146	0.3739	0.8454
0.4026	8.99	164	0.4483	0.8249
0.4326	9.97	182	0.3992	0.8447
0.4149	10.96	200	0.3607	0.8542
0.3995	12.0	219	0.4662	0.8256
0.36	12.99	237	0.4375	0.8495
0.3807	13.97	255	0.4013	0.8351
0.401	14.96	273	0.4875	0.8311
0.3349	16.0	292	0.3810	0.8610
0.3279	16.99	310	0.4288	0.8392
0.3111	17.97	328	0.4160	0.8460
0.3092	18.96	346	0.4469	0.8379
0.3202	20.0	365	0.4294	0.8563
0.3027	20.99	383	0.3928	0.8569
0.3022	21.97	401	0.4829	0.8399
0.2934	22.96	419	0.3978	0.8604
0.2789	24.0	438	0.4027	0.8610
0.2714	24.66	450	0.4265	0.8569

Framework versions

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