wav2vec2-large-robust-ft-libri-960h-finetuned-ravdess-v2

This model is a fine-tuned version of facebook/wav2vec2-large-robust-ft-libri-960h on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0280
• Accuracy: 0.6146

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• 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
2.0786	1.0	36	2.0692	0.1597
2.0578	2.0	72	2.0555	0.1979
1.9903	3.0	108	1.9172	0.2882
1.8052	4.0	144	1.7975	0.2951
1.7221	5.0	180	1.6602	0.4028
1.5773	6.0	216	1.6362	0.4479
1.4785	7.0	252	1.4675	0.4965
1.3828	8.0	288	1.3735	0.5
1.2352	9.0	324	1.2886	0.5278
1.159	10.0	360	1.2184	0.5521
1.073	11.0	396	1.1456	0.5556
1.0127	12.0	432	1.1864	0.5694
0.9374	13.0	468	1.1865	0.5625
0.8622	14.0	504	1.1745	0.5660
0.8704	15.0	540	1.1563	0.5694
0.8607	16.0	576	1.0466	0.5938
0.8228	17.0	612	1.0457	0.6007
0.8521	18.0	648	1.0280	0.6146
0.8248	19.0	684	1.0399	0.6146
0.7901	20.0	720	1.0402	0.6111

Framework versions

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