model_syllable_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: 0.1349
• 0 Precision: 1.0
• 0 Recall: 1.0
• 0 F1-score: 1.0
• 0 Support: 26
• 1 Precision: 1.0
• 1 Recall: 0.9677
• 1 F1-score: 0.9836
• 1 Support: 31
• 2 Precision: 0.9630
• 2 Recall: 1.0
• 2 F1-score: 0.9811
• 2 Support: 26
• 3 Precision: 1.0
• 3 Recall: 1.0
• 3 F1-score: 1.0
• 3 Support: 14
• Accuracy: 0.9897
• Macro avg Precision: 0.9907
• Macro avg Recall: 0.9919
• Macro avg F1-score: 0.9912
• Macro avg Support: 97
• Weighted avg Precision: 0.9901
• Weighted avg Recall: 0.9897
• Weighted avg F1-score: 0.9897
• Weighted avg Support: 97
• Wer: 0.2258
• Mtrix: [[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 30, 1, 0], [2, 0, 0, 26, 0], [3, 0, 0, 0, 14]]

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: 70
• 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
1.6602	4.16	100	1.5639	0.0	0.0	0.0	26	0.0	0.0	0.0	31	0.2584	0.8846	0.4	26	0.0	0.0	0.0	14	0.2371	0.0646	0.2212	0.1	97	0.0693	0.2371	0.1072	97	0.9732	[[0, 1, 2, 3], [0, 0, 0, 26, 0], [1, 0, 0, 31, 0], [2, 3, 0, 23, 0], [3, 5, 0, 9, 0]]
1.616	8.33	200	1.4203	0.0	0.0	0.0	26	0.0	0.0	0.0	31	0.2584	0.8846	0.4	26	0.0	0.0	0.0	14	0.2371	0.0646	0.2212	0.1	97	0.0693	0.2371	0.1072	97	0.9732	[[0, 1, 2, 3], [0, 0, 0, 26, 0], [1, 0, 0, 31, 0], [2, 3, 0, 23, 0], [3, 5, 0, 9, 0]]
1.2107	12.49	300	1.1249	0.0	0.0	0.0	26	0.0	0.0	0.0	31	0.2584	0.8846	0.4	26	0.0	0.0	0.0	14	0.2371	0.0646	0.2212	0.1	97	0.0693	0.2371	0.1072	97	0.9732	[[0, 1, 2, 3], [0, 0, 0, 26, 0], [1, 0, 0, 31, 0], [2, 3, 0, 23, 0], [3, 5, 0, 9, 0]]
1.1283	16.65	400	1.0201	0.0	0.0	0.0	26	0.0	0.0	0.0	31	0.2584	0.8846	0.4	26	0.0	0.0	0.0	14	0.2371	0.0646	0.2212	0.1	97	0.0693	0.2371	0.1072	97	0.9732	[[0, 1, 2, 3], [0, 0, 0, 26, 0], [1, 0, 0, 31, 0], [2, 3, 0, 23, 0], [3, 5, 0, 9, 0]]
0.8868	20.82	500	0.8944	0.0	0.0	0.0	26	0.0	0.0	0.0	31	0.2584	0.8846	0.4	26	0.0	0.0	0.0	14	0.2371	0.0646	0.2212	0.1	97	0.0693	0.2371	0.1072	97	0.9732	[[0, 1, 2, 3], [0, 0, 0, 26, 0], [1, 0, 0, 31, 0], [2, 3, 0, 23, 0], [3, 5, 0, 9, 0]]
0.8863	24.98	600	0.9316	0.0	0.0	0.0	26	0.0	0.0	0.0	31	0.2584	0.8846	0.4	26	0.0	0.0	0.0	14	0.2371	0.0646	0.2212	0.1	97	0.0693	0.2371	0.1072	97	0.9732	[[0, 1, 2, 3], [0, 0, 0, 26, 0], [1, 0, 0, 31, 0], [2, 3, 0, 23, 0], [3, 5, 0, 9, 0]]
0.9019	29.16	700	0.8688	0.7647	1.0	0.8667	26	0.0	0.0	0.0	31	0.3651	0.8846	0.5169	26	0.0	0.0	0.0	14	0.5052	0.2824	0.4712	0.3459	97	0.3028	0.5052	0.3708	97	0.9732	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 0, 31, 0], [2, 3, 0, 23, 0], [3, 5, 0, 9, 0]]
0.7977	33.33	800	0.8014	1.0	1.0	1.0	26	0.9667	0.9355	0.9508	31	0.9259	0.9615	0.9434	26	1.0	1.0	1.0	14	0.9691	0.9731	0.9743	0.9736	97	0.9695	0.9691	0.9691	97	1.0	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 29, 2, 0], [2, 0, 1, 25, 0], [3, 0, 0, 0, 14]]
0.729	37.49	900	0.8163	1.0	1.0	1.0	26	0.9091	0.9677	0.9375	31	0.9583	0.8846	0.9200	26	1.0	1.0	1.0	14	0.9588	0.9669	0.9631	0.9644	97	0.9598	0.9588	0.9586	97	1.0	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 30, 1, 0], [2, 0, 3, 23, 0], [3, 0, 0, 0, 14]]
0.6526	41.65	1000	0.6691	1.0	1.0	1.0	26	0.9667	0.9355	0.9508	31	0.9259	0.9615	0.9434	26	1.0	1.0	1.0	14	0.9691	0.9731	0.9743	0.9736	97	0.9695	0.9691	0.9691	97	0.7055	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 29, 2, 0], [2, 0, 1, 25, 0], [3, 0, 0, 0, 14]]
0.6633	45.82	1100	0.3445	1.0	1.0	1.0	26	0.9394	1.0	0.9688	31	1.0	0.9231	0.9600	26	1.0	1.0	1.0	14	0.9794	0.9848	0.9808	0.9822	97	0.9806	0.9794	0.9793	97	0.5017	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 31, 0, 0], [2, 0, 2, 24, 0], [3, 0, 0, 0, 14]]
0.1913	49.98	1200	0.2455	1.0	1.0	1.0	26	0.9677	0.9677	0.9677	31	0.96	0.9231	0.9412	26	0.9333	1.0	0.9655	14	0.9691	0.9653	0.9727	0.9686	97	0.9693	0.9691	0.9689	97	0.3946	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 30, 1, 0], [2, 0, 1, 24, 1], [3, 0, 0, 0, 14]]
0.2024	54.16	1300	0.1865	1.0	1.0	1.0	26	1.0	0.9355	0.9667	31	0.9286	1.0	0.9630	26	1.0	1.0	1.0	14	0.9794	0.9821	0.9839	0.9824	97	0.9809	0.9794	0.9794	97	0.3423	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 29, 2, 0], [2, 0, 0, 26, 0], [3, 0, 0, 0, 14]]
0.1212	58.33	1400	0.1485	1.0	1.0	1.0	26	1.0	0.9677	0.9836	31	0.9630	1.0	0.9811	26	1.0	1.0	1.0	14	0.9897	0.9907	0.9919	0.9912	97	0.9901	0.9897	0.9897	97	0.2957	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 30, 1, 0], [2, 0, 0, 26, 0], [3, 0, 0, 0, 14]]
0.108	62.49	1500	0.1348	1.0	1.0	1.0	26	1.0	0.9677	0.9836	31	0.9630	1.0	0.9811	26	1.0	1.0	1.0	14	0.9897	0.9907	0.9919	0.9912	97	0.9901	0.9897	0.9897	97	0.2433	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 30, 1, 0], [2, 0, 0, 26, 0], [3, 0, 0, 0, 14]]
0.1058	66.65	1600	0.1328	1.0	1.0	1.0	26	1.0	0.9677	0.9836	31	0.9630	1.0	0.9811	26	1.0	1.0	1.0	14	0.9897	0.9907	0.9919	0.9912	97	0.9901	0.9897	0.9897	97	0.2224	[[0, 1, 2, 3], [0, 26, 0, 0, 0], [1, 0, 30, 1, 0], [2, 0, 0, 26, 0], [3, 0, 0, 0, 14]]

Framework versions

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