model_broadclass_onSet4

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1340
• 0 Precision: 1.0
• 0 Recall: 0.9615
• 0 F1-score: 0.9804
• 0 Support: 26
• 1 Precision: 1.0
• 1 Recall: 1.0
• 1 F1-score: 1.0
• 1 Support: 32
• 2 Precision: 1.0
• 2 Recall: 0.9643
• 2 F1-score: 0.9818
• 2 Support: 28
• 3 Precision: 0.8462
• 3 Recall: 1.0
• 3 F1-score: 0.9167
• 3 Support: 11
• Accuracy: 0.9794
• Macro avg Precision: 0.9615
• Macro avg Recall: 0.9815
• Macro avg F1-score: 0.9697
• Macro avg Support: 97
• Weighted avg Precision: 0.9826
• Weighted avg Recall: 0.9794
• Weighted avg F1-score: 0.9800
• Weighted avg Support: 97
• Wer: 0.1098
• Mtrix: [[0, 1, 2, 3], [0, 25, 0, 0, 1], [1, 0, 32, 0, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 11]]

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: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 80
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	0 Precision	0 Recall	0 F1-score	0 Support	1 Precision	1 Recall	1 F1-score	1 Support	2 Precision	2 Recall	2 F1-score	2 Support	3 Precision	3 Recall	3 F1-score	3 Support	Accuracy	Macro avg Precision	Macro avg Recall	Macro avg F1-score	Macro avg Support	Weighted avg Precision	Weighted avg Recall	Weighted avg F1-score	Weighted avg Support	Wer	Mtrix
2.337	4.16	100	2.1761	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9869	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
2.2604	8.33	200	2.0783	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9869	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
1.9239	12.49	300	1.8395	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9869	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
1.7002	16.65	400	1.7194	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9869	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
1.611	20.82	500	1.5619	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9869	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
1.486	24.98	600	1.5283	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9869	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
1.6085	29.16	700	1.5041	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9869	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
1.5607	33.33	800	1.4456	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9945	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
1.3499	37.49	900	1.2898	0.2680	1.0	0.4228	26	0.0	0.0	0.0	32	0.0	0.0	0.0	28	0.0	0.0	0.0	11	0.2680	0.0670	0.25	0.1057	97	0.0718	0.2680	0.1133	97	0.9970	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 32, 0, 0, 0], [2, 28, 0, 0, 0], [3, 11, 0, 0, 0]]
0.9722	41.65	1000	0.9757	0.3133	1.0	0.4771	26	1.0	0.1562	0.2703	32	1.0	0.1786	0.3030	28	1.0	0.3636	0.5333	11	0.4124	0.8283	0.4246	0.3959	97	0.8159	0.4124	0.3650	97	0.9612	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 27, 5, 0, 0], [2, 23, 0, 5, 0], [3, 7, 0, 0, 4]]
0.9679	45.82	1100	0.9452	0.4333	1.0	0.6047	26	0.9630	0.8125	0.8814	32	1.0	0.3214	0.4865	28	1.0	0.0909	0.1667	11	0.6392	0.8491	0.5562	0.5348	97	0.8359	0.6392	0.6122	97	0.9406	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 6, 26, 0, 0], [2, 18, 1, 9, 0], [3, 10, 0, 0, 1]]
0.9206	49.98	1200	0.9031	0.5909	1.0	0.7429	26	1.0	0.9062	0.9508	32	1.0	0.7143	0.8333	28	1.0	0.3636	0.5333	11	0.8144	0.8977	0.7460	0.7651	97	0.8903	0.8144	0.8138	97	0.9250	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 3, 29, 0, 0], [2, 8, 0, 20, 0], [3, 7, 0, 0, 4]]
0.9223	54.16	1300	0.8607	0.8125	1.0	0.8966	26	1.0	0.875	0.9333	32	1.0	0.9643	0.9818	28	1.0	0.9091	0.9524	11	0.9381	0.9531	0.9371	0.9410	97	0.9497	0.9381	0.9396	97	0.9366	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 4, 28, 0, 0], [2, 1, 0, 27, 0], [3, 1, 0, 0, 10]]
0.8407	58.33	1400	0.8011	0.8929	0.9615	0.9259	26	1.0	0.9688	0.9841	32	1.0	0.8929	0.9434	28	0.8462	1.0	0.9167	11	0.9485	0.9348	0.9558	0.9425	97	0.9538	0.9485	0.9491	97	0.9381	[[0, 1, 2, 3], [0, 25, 0, 0, 1], [1, 1, 31, 0, 0], [2, 2, 0, 25, 1], [3, 0, 0, 0, 11]]
0.7359	62.49	1500	0.7210	0.8966	1.0	0.9455	26	1.0	0.9375	0.9677	32	1.0	0.9286	0.9630	28	0.9167	1.0	0.9565	11	0.9588	0.9533	0.9665	0.9582	97	0.9628	0.9588	0.9591	97	0.9220	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 2, 30, 0, 0], [2, 1, 0, 26, 1], [3, 0, 0, 0, 11]]
0.5479	66.65	1600	0.4813	1.0	1.0	1.0	26	1.0	1.0	1.0	32	1.0	0.9643	0.9818	28	0.9167	1.0	0.9565	11	0.9897	0.9792	0.9911	0.9846	97	0.9905	0.9897	0.9898	97	0.7447	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 32, 0, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 11]]
0.2617	70.82	1700	0.2138	1.0	1.0	1.0	26	1.0	1.0	1.0	32	1.0	0.9643	0.9818	28	0.9167	1.0	0.9565	11	0.9897	0.9792	0.9911	0.9846	97	0.9905	0.9897	0.9898	97	0.1692	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 32, 0, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 11]]
0.2186	74.98	1800	0.1412	1.0	1.0	1.0	26	1.0	1.0	1.0	32	1.0	0.9643	0.9818	28	0.9167	1.0	0.9565	11	0.9897	0.9792	0.9911	0.9846	97	0.9905	0.9897	0.9898	97	0.1269	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 32, 0, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 11]]
0.2303	79.16	1900	0.1344	1.0	0.9615	0.9804	26	1.0	1.0	1.0	32	1.0	0.9643	0.9818	28	0.8462	1.0	0.9167	11	0.9794	0.9615	0.9815	0.9697	97	0.9826	0.9794	0.9800	97	0.1113	[[0, 1, 2, 3], [0, 25, 0, 0, 1], [1, 0, 32, 0, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 11]]

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu116
• Datasets 2.8.0
• Tokenizers 0.13.2