my_awesome_emotions_model

This model is a fine-tuned version of facebook/wav2vec2-base on the audiofolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.3229
• Accuracy: 0.6086

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.7077	0.95	14	2.7049	0.0645
2.7036	1.97	29	2.6957	0.0688
2.6761	2.98	44	2.6471	0.1075
2.6414	4.0	59	2.5442	0.1398
2.5034	4.95	73	2.3988	0.2774
2.4337	5.97	88	2.2798	0.3097
2.2757	6.98	103	2.2054	0.3290
2.2346	8.0	118	2.1204	0.3484
2.1167	8.95	132	2.0784	0.3484
2.0801	9.97	147	2.0267	0.3484
1.987	10.98	162	1.9672	0.3699
1.9278	12.0	177	1.9686	0.3505
1.839	12.95	191	1.8604	0.4215
1.785	13.97	206	1.8611	0.4129
1.6811	14.98	221	1.7791	0.4473
1.6858	16.0	236	1.7533	0.4323
1.562	16.95	250	1.7364	0.4473
1.5469	17.97	265	1.7407	0.4430
1.485	18.98	280	1.7055	0.4301
1.4489	20.0	295	1.6566	0.4839
1.426	20.95	309	1.5844	0.5054
1.3596	21.97	324	1.6252	0.4796
1.3212	22.98	339	1.5797	0.4860
1.248	24.0	354	1.5483	0.5097
1.1954	24.95	368	1.5301	0.5290
1.1629	25.97	383	1.4905	0.5398
1.1364	26.98	398	1.5040	0.5355
1.0897	28.0	413	1.5128	0.5484
1.0564	28.95	427	1.4761	0.5570
1.0149	29.97	442	1.4948	0.5247
0.975	30.98	457	1.4194	0.5742
0.9546	32.0	472	1.3986	0.5763
0.9235	32.95	486	1.4126	0.5634
0.8848	33.97	501	1.4284	0.5763
0.8579	34.98	516	1.3872	0.5677
0.8475	36.0	531	1.4108	0.5742
0.8018	36.95	545	1.3667	0.5849
0.7861	37.97	560	1.3614	0.5914
0.7756	38.98	575	1.3473	0.5914
0.7427	40.0	590	1.3346	0.5914
0.764	40.95	604	1.3229	0.6086
0.7283	41.97	619	1.3206	0.6
0.714	42.98	634	1.3266	0.6065
0.724	44.0	649	1.3377	0.5957
0.6928	44.95	663	1.3281	0.5978
0.7065	45.97	678	1.3214	0.6086
0.6781	46.98	693	1.3338	0.5849
0.7058	47.46	700	1.3330	0.5892

Framework versions

• Transformers 4.36.0
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0