hubert-large-ll60k-finetuned-ravdess-v4

This model is a fine-tuned version of facebook/hubert-large-ll60k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0831
• Accuracy: 0.625

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
2.0806	1.0	18	2.0768	0.1493
2.0707	2.0	36	2.0653	0.1771
2.0551	3.0	54	2.0475	0.1771
2.0223	4.0	72	2.0066	0.2153
1.8973	5.0	90	1.8130	0.2882
1.8547	6.0	108	1.8352	0.2847
1.7718	7.0	126	1.8253	0.2847
1.722	8.0	144	1.7302	0.2986
1.7004	9.0	162	1.7978	0.3160
1.6574	10.0	180	1.6873	0.3507
1.6387	11.0	198	1.7008	0.3576
1.5605	12.0	216	1.5569	0.4340
1.488	13.0	234	1.4882	0.4444
1.4243	14.0	252	1.4232	0.4653
1.3775	15.0	270	1.4478	0.4410
1.3716	16.0	288	1.3412	0.5035
1.2768	17.0	306	1.3833	0.4688
1.2841	18.0	324	1.3048	0.5069
1.2193	19.0	342	1.2395	0.5521
1.202	20.0	360	1.2025	0.5660
1.169	21.0	378	1.2574	0.5312
1.164	22.0	396	1.1877	0.5521
1.0749	23.0	414	1.1508	0.5764
1.0779	24.0	432	1.1783	0.5556
1.0732	25.0	450	1.1438	0.5625
1.016	26.0	468	1.1971	0.5590
1.022	27.0	486	1.1014	0.6042
1.0345	28.0	504	1.0626	0.6215
1.0081	29.0	522	1.0928	0.6181
1.0036	30.0	540	1.0831	0.625

Framework versions

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