Wav2vec2-xlsr-Shemo-Ravdess-4EMO

This model is a fine-tuned version of makhataei/Wav2vec2-xlsr-Shemo-Ravdess-4EMO on the minoosh/shEMO dataset. It achieves the following results on the evaluation set:

• Loss: 0.7652
• Accuracy: 0.7256

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: 0.0005
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• 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
0.6733	1.0	250	0.7858	0.7256
0.6866	2.0	500	0.7594	0.7143
0.6558	3.0	750	0.8044	0.7279
0.6496	4.0	1000	0.7633	0.7188
0.6393	5.0	1250	0.7901	0.7098
0.6319	6.0	1500	0.7739	0.7075
0.626	7.0	1750	0.7509	0.7302
0.6335	8.0	2000	0.7741	0.7166
0.6376	9.0	2250	0.7856	0.7143
0.6621	10.0	2500	0.7515	0.7188
0.6379	11.0	2750	0.7932	0.7166
0.6186	12.0	3000	0.7728	0.7166
0.6189	13.0	3250	0.7640	0.7143
0.6215	14.0	3500	0.7579	0.7211
0.6187	15.0	3750	0.7822	0.7166
0.6267	16.0	4000	0.7863	0.7143
0.6144	17.0	4250	0.7654	0.7188
0.6138	18.0	4500	0.7719	0.7166
0.5821	19.0	4750	0.7707	0.7234
0.5835	20.0	5000	0.7652	0.7256

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu117
• Datasets 2.12.0
• Tokenizers 0.13.3