wav2vec2-large-robust-ft-libri-960h-finetuned-ravdess-v3

This model is a fine-tuned version of facebook/wav2vec2-large-robust-ft-libri-960h on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9736
• Accuracy: 0.6354

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0795	1.0	18	2.0783	0.1007
2.0703	2.0	36	2.0743	0.1181
2.062	3.0	54	2.0632	0.1597
2.0444	4.0	72	2.0439	0.1910
2.0031	5.0	90	1.9762	0.2778
1.9632	6.0	108	1.8421	0.3576
1.8249	7.0	126	1.7072	0.3889
1.6733	8.0	144	1.5729	0.3819
1.5452	9.0	162	1.4826	0.4201
1.4479	10.0	180	1.4214	0.4236
1.443	11.0	198	1.3441	0.4340
1.3341	12.0	216	1.3241	0.5
1.2697	13.0	234	1.2810	0.5069
1.2348	14.0	252	1.2349	0.5069
1.1785	15.0	270	1.1948	0.5208
1.1687	16.0	288	1.1831	0.5451
1.1168	17.0	306	1.1481	0.5764
1.0975	18.0	324	1.1342	0.5764
1.0491	19.0	342	1.1138	0.6146
1.033	20.0	360	1.0800	0.6146
1.0523	21.0	378	1.0678	0.6146
1.0136	22.0	396	1.0472	0.6111
0.9777	23.0	414	1.0175	0.6111
1.0007	24.0	432	1.0703	0.6215
0.9584	25.0	450	0.9935	0.6181
0.9102	26.0	468	0.9736	0.6354
0.9101	27.0	486	0.9758	0.6285
0.9405	28.0	504	0.9659	0.6319
0.9366	29.0	522	0.9719	0.625
0.9498	30.0	540	0.9713	0.6215

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3