wav2vec2-large-finetuned-iemocap2

This model is a fine-tuned version of facebook/wav2vec2-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2460
• Accuracy: 0.5209
• F1: 0.5049

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.3574	0.98	25	1.4694	0.2502	0.1002
1.1919	1.98	50	1.3444	0.3754	0.3304
1.1571	2.98	75	1.2644	0.4064	0.3649
1.1165	3.98	100	1.1895	0.4762	0.4223
1.0498	4.98	125	1.1373	0.5053	0.4920
1.0147	5.98	150	1.1089	0.5131	0.4763
1.0163	6.98	175	1.1595	0.5092	0.4651
0.9711	7.98	200	1.1298	0.5179	0.4759
0.9599	8.98	225	1.1460	0.5199	0.4831
0.9042	9.98	250	1.1191	0.5500	0.5307
0.8734	10.98	275	1.2103	0.5364	0.4935
0.8876	11.98	300	1.1837	0.5228	0.4912
0.8369	12.98	325	1.2009	0.5296	0.4927
0.8357	13.98	350	1.2144	0.5238	0.5054
0.8314	14.98	375	1.1866	0.5335	0.5180
0.761	15.98	400	1.2145	0.5451	0.5317
0.7723	16.98	425	1.2033	0.5276	0.5073
0.7775	17.98	450	1.2841	0.5228	0.4986
0.7735	18.98	475	1.2249	0.5393	0.5253
0.726	19.98	500	1.2460	0.5209	0.5049

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2