speech-emotion-recognition-wav2vec2

This model is a fine-tuned version of jonatasgrosman/wav2vec2-large-xlsr-53-english on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2842
• Accuracy: 0.9045

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: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.1026	0.0236	10	2.0265	0.1592
1.9631	0.0472	20	2.0125	0.1993
1.9106	0.0708	30	1.8609	0.2417
1.715	0.0943	40	1.7659	0.3054
1.69	0.1179	50	1.5524	0.3785
1.4684	0.1415	60	1.4516	0.4057
1.3422	0.1651	70	1.2702	0.5354
1.2358	0.1887	80	0.9599	0.6899
0.9937	0.2123	90	0.8447	0.7394
0.7604	0.2358	100	0.8068	0.7453
0.7736	0.2594	110	0.6561	0.7913
0.6573	0.2830	120	0.6584	0.7830
0.5634	0.3066	130	0.5564	0.8066
0.5353	0.3302	140	0.5586	0.8184
0.3805	0.3538	150	0.6575	0.7818
0.6584	0.3774	160	0.4686	0.8538
0.4788	0.4009	170	0.4533	0.8514
0.4123	0.4245	180	0.5266	0.8432
0.4964	0.4481	190	0.5038	0.8325
0.4489	0.4717	200	0.5552	0.8208
0.4562	0.4953	210	0.4075	0.8526
0.5362	0.5189	220	0.4975	0.8184
0.3539	0.5425	230	0.4947	0.8267
0.4726	0.5660	240	0.4456	0.8514
0.3897	0.5896	250	0.3567	0.8715
0.2817	0.6132	260	0.3880	0.8644
0.3281	0.6368	270	0.3902	0.8679
0.311	0.6604	280	0.3243	0.9021
0.1768	0.6840	290	0.4162	0.8644
0.3748	0.7075	300	0.4482	0.8644
0.588	0.7311	310	0.3179	0.8950
0.402	0.7547	320	0.2955	0.9033
0.4068	0.7783	330	0.3212	0.8962
0.3622	0.8019	340	0.3931	0.8550
0.4407	0.8255	350	0.3467	0.8644
0.3474	0.8491	360	0.3149	0.8962
0.3449	0.8726	370	0.2829	0.9033
0.2673	0.8962	380	0.2566	0.9198
0.2998	0.9198	390	0.2614	0.9127
0.2721	0.9434	400	0.2786	0.9021
0.2717	0.9670	410	0.2891	0.9021
0.3277	0.9906	420	0.2842	0.9045

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1