id_ravdess_mel_spec_Vit_swin-tiny-patch4-window7-224_1

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.3684
Accuracy: 0.8935

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: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 128
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	8	3.1637	0.0509
3.1954	2.0	16	3.0952	0.1343
3.1006	3.0	24	2.9019	0.2083
2.8528	4.0	32	2.5496	0.3611
2.3055	5.0	40	1.9709	0.4583
2.3055	6.0	48	1.4990	0.5880
1.5571	7.0	56	1.2140	0.6898
1.0455	8.0	64	1.0802	0.6574
0.7066	9.0	72	0.8690	0.75
0.4544	10.0	80	0.6908	0.7778
0.4544	11.0	88	0.5625	0.8102
0.2632	12.0	96	0.5499	0.8194
0.1741	13.0	104	0.5504	0.8056
0.1183	14.0	112	0.5433	0.8056
0.0809	15.0	120	0.5394	0.8194
0.0809	16.0	128	0.4215	0.8565
0.0663	17.0	136	0.4608	0.8704
0.0487	18.0	144	0.4469	0.8426
0.0433	19.0	152	0.4972	0.8519
0.0366	20.0	160	0.3546	0.8796
0.0366	21.0	168	0.4419	0.8657
0.0193	22.0	176	0.5580	0.8519
0.0192	23.0	184	0.5105	0.8426
0.0234	24.0	192	0.3844	0.8611
0.017	25.0	200	0.4979	0.875
0.017	26.0	208	0.4034	0.8657
0.0194	27.0	216	0.4212	0.8889
0.0168	28.0	224	0.4298	0.8843
0.0217	29.0	232	0.3586	0.8981
0.0097	30.0	240	0.4900	0.8796
0.0097	31.0	248	0.4183	0.8565
0.0057	32.0	256	0.7694	0.8380
0.0133	33.0	264	0.4215	0.8657
0.0198	34.0	272	0.5039	0.8657
0.0113	35.0	280	0.4496	0.8796
0.0113	36.0	288	0.4230	0.8889
0.0026	37.0	296	0.5184	0.8565
0.0095	38.0	304	0.4038	0.8889
0.0059	39.0	312	0.4309	0.875
0.0097	40.0	320	0.6481	0.8148
0.0097	41.0	328	0.3568	0.8935
0.0058	42.0	336	0.3418	0.9259
0.0035	43.0	344	0.3504	0.9213
0.0042	44.0	352	0.3305	0.9167
0.0018	45.0	360	0.3697	0.9028
0.0018	46.0	368	0.3399	0.9028
0.0049	47.0	376	0.3395	0.9028
0.002	48.0	384	0.3574	0.9028
0.0031	49.0	392	0.3696	0.8981
0.0011	50.0	400	0.3684	0.8935

Framework versions

Transformers 4.47.0
Pytorch 2.5.1+cu121
Datasets 3.3.1
Tokenizers 0.21.0

Loading…

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-05

train_batch_size: 32

eval_batch_size: 32

seed: 42

gradient_accumulation_steps: 4

total_train_batch_size: 128

optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments

lr_scheduler_type: linear

lr_scheduler_warmup_ratio: 0.1

num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	8	3.1637	0.0509
3.1954	2.0	16	3.0952	0.1343
3.1006	3.0	24	2.9019	0.2083
2.8528	4.0	32	2.5496	0.3611
2.3055	5.0	40	1.9709	0.4583
2.3055	6.0	48	1.4990	0.5880
1.5571	7.0	56	1.2140	0.6898
1.0455	8.0	64	1.0802	0.6574
0.7066	9.0	72	0.8690	0.75
0.4544	10.0	80	0.6908	0.7778
0.4544	11.0	88	0.5625	0.8102
0.2632	12.0	96	0.5499	0.8194
0.1741	13.0	104	0.5504	0.8056
0.1183	14.0	112	0.5433	0.8056
0.0809	15.0	120	0.5394	0.8194
0.0809	16.0	128	0.4215	0.8565
0.0663	17.0	136	0.4608	0.8704
0.0487	18.0	144	0.4469	0.8426
0.0433	19.0	152	0.4972	0.8519
0.0366	20.0	160	0.3546	0.8796
0.0366	21.0	168	0.4419	0.8657
0.0193	22.0	176	0.5580	0.8519
0.0192	23.0	184	0.5105	0.8426
0.0234	24.0	192	0.3844	0.8611
0.017	25.0	200	0.4979	0.875
0.017	26.0	208	0.4034	0.8657
0.0194	27.0	216	0.4212	0.8889
0.0168	28.0	224	0.4298	0.8843
0.0217	29.0	232	0.3586	0.8981
0.0097	30.0	240	0.4900	0.8796
0.0097	31.0	248	0.4183	0.8565
0.0057	32.0	256	0.7694	0.8380
0.0133	33.0	264	0.4215	0.8657
0.0198	34.0	272	0.5039	0.8657
0.0113	35.0	280	0.4496	0.8796
0.0113	36.0	288	0.4230	0.8889
0.0026	37.0	296	0.5184	0.8565
0.0095	38.0	304	0.4038	0.8889
0.0059	39.0	312	0.4309	0.875
0.0097	40.0	320	0.6481	0.8148
0.0097	41.0	328	0.3568	0.8935
0.0058	42.0	336	0.3418	0.9259
0.0035	43.0	344	0.3504	0.9213
0.0042	44.0	352	0.3305	0.9167
0.0018	45.0	360	0.3697	0.9028
0.0018	46.0	368	0.3399	0.9028
0.0049	47.0	376	0.3395	0.9028
0.002	48.0	384	0.3574	0.9028
0.0031	49.0	392	0.3696	0.8981
0.0011	50.0	400	0.3684	0.8935

Framework versions

Transformers 4.47.0

Pytorch 2.5.1+cu121

Datasets 3.3.1

Tokenizers 0.21.0

ricardoSLabs/id_ravdess_mel_spec_Vit_swin-tiny-patch4-window7-224_1

id_ravdess_mel_spec_Vit_swin-tiny-patch4-window7-224_1

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

id_ravdess_mel_spec_Vit_swin-tiny-patch4-window7-224_1

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions