wavlm-large-finetuned-iemocap2

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

• Loss: 1.0935
• Accuracy: 0.5335
• F1: 0.5005

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
1.3826	0.98	25	1.3815	0.2502	0.1003
1.3263	1.98	50	1.3663	0.2502	0.1002
1.2563	2.98	75	1.2589	0.3870	0.3051
1.1869	3.98	100	1.2042	0.3977	0.3428
1.1291	4.98	125	1.1768	0.4539	0.4557
1.1171	5.98	150	1.1425	0.4888	0.4799
1.0811	6.98	175	1.1316	0.4956	0.4851
1.0627	7.98	200	1.1241	0.5044	0.4859
1.079	8.98	225	1.1026	0.5228	0.5031
1.0294	9.98	250	1.1018	0.5199	0.4959
1.0088	10.98	275	1.0903	0.5325	0.5046
1.0217	11.98	300	1.0966	0.5296	0.5015
1.0034	12.98	325	1.1012	0.5296	0.4990
1.0024	13.98	350	1.0832	0.5393	0.5127
1.0047	14.98	375	1.0902	0.5315	0.4986
0.9436	15.98	400	1.0896	0.5373	0.5085
0.9584	16.98	425	1.0859	0.5412	0.5114
0.9859	17.98	450	1.0865	0.5412	0.5120
0.9679	18.98	475	1.0926	0.5335	0.4999
0.9468	19.98	500	1.0935	0.5335	0.5005

Framework versions

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