adhityamw11/distilhubert-finetuned_distillhubert-ravdess

distilhubert-finetuned-ravdess

This model is a fine-tuned version of ntu-spml/distilhubert on the RAVDESS dataset. It achieves the following results on the evaluation set:

• Loss: 0.9331
• Accuracy: 0.8438

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: 8
• eval_batch_size: 8
• seed: 42
• 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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0641	1.0	144	2.0414	0.2778
1.751	2.0	288	1.7801	0.3854
1.5345	3.0	432	1.3610	0.5417
1.1913	4.0	576	1.1896	0.5417
0.8227	5.0	720	0.7924	0.7535
0.6563	6.0	864	0.6772	0.7743
0.4082	7.0	1008	0.6398	0.7847
0.5133	8.0	1152	0.6409	0.7951
0.0467	9.0	1296	0.7356	0.7951
0.0232	10.0	1440	0.8220	0.8160
0.0298	11.0	1584	0.7164	0.8438
0.0021	12.0	1728	0.7578	0.8611
0.0014	13.0	1872	0.6806	0.8507
0.0012	14.0	2016	0.6953	0.8507
0.0009	15.0	2160	0.7311	0.8403
0.0007	16.0	2304	0.7312	0.8472
0.0006	17.0	2448	0.7528	0.8438
0.0005	18.0	2592	0.7748	0.8299
0.0005	19.0	2736	0.7692	0.8472
0.0004	20.0	2880	0.7806	0.8403
0.0003	21.0	3024	0.7907	0.8438
0.0003	22.0	3168	0.7909	0.8438
0.0003	23.0	3312	0.8060	0.8472
0.0003	24.0	3456	0.8302	0.8438
0.0002	25.0	3600	0.8296	0.8438
0.0002	26.0	3744	0.8306	0.8403
0.0002	27.0	3888	0.8399	0.8438
0.0002	28.0	4032	0.8447	0.8438
0.0002	29.0	4176	0.8488	0.8403
0.0002	30.0	4320	0.8564	0.8472
0.0002	31.0	4464	0.8618	0.8472
0.0001	32.0	4608	0.8736	0.8438
0.0001	33.0	4752	0.8793	0.8403
0.0001	34.0	4896	0.8840	0.8438
0.0001	35.0	5040	0.8870	0.8438
0.0001	36.0	5184	0.8882	0.8472
0.0001	37.0	5328	0.9033	0.8403
0.0001	38.0	5472	0.8980	0.8403
0.0001	39.0	5616	0.9081	0.8472
0.0001	40.0	5760	0.9086	0.8472
0.0001	41.0	5904	0.9119	0.8438
0.0001	42.0	6048	0.9106	0.8507
0.0001	43.0	6192	0.9188	0.8438
0.0001	44.0	6336	0.9238	0.8438
0.0001	45.0	6480	0.9282	0.8438
0.0001	46.0	6624	0.9286	0.8438
0.0001	47.0	6768	0.9312	0.8438
0.0001	48.0	6912	0.9296	0.8472
0.0001	49.0	7056	0.9324	0.8438
0.0001	50.0	7200	0.9331	0.8438

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.2