disfluency-large-3

This model is a fine-tuned version of vinai/phobert-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 42.3155
• Precision: 0.9886
• Recall: 0.9862
• F1: 0.9874
• Accuracy: 0.9956

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	140	131.9614	0.8037	0.8438	0.8232	0.9411
No log	2.0	280	41.8031	0.9487	0.9549	0.9518	0.9855
No log	3.0	420	27.3502	0.9664	0.9681	0.9673	0.9906
178.6738	4.0	560	22.9255	0.9741	0.9730	0.9735	0.9925
178.6738	5.0	700	25.3163	0.9676	0.9688	0.9682	0.9919
178.6738	6.0	840	24.1142	0.9723	0.9718	0.9720	0.9925
178.6738	7.0	980	22.2517	0.9777	0.9766	0.9771	0.9938
25.7318	8.0	1120	24.4542	0.9760	0.9772	0.9766	0.9936
25.7318	9.0	1260	27.1333	0.9740	0.9700	0.9720	0.9929
25.7318	10.0	1400	24.5889	0.9789	0.9778	0.9784	0.9938
16.0059	11.0	1540	26.1038	0.9819	0.9808	0.9814	0.9936
16.0059	12.0	1680	23.3198	0.9790	0.9814	0.9802	0.9941
16.0059	13.0	1820	30.8831	0.9778	0.9772	0.9775	0.9930
16.0059	14.0	1960	28.1502	0.9843	0.9814	0.9828	0.9946
11.2302	15.0	2100	29.2842	0.9790	0.9808	0.9799	0.9937
11.2302	16.0	2240	28.5446	0.9819	0.9796	0.9807	0.9945
11.2302	17.0	2380	25.4603	0.9850	0.9838	0.9844	0.9953
8.9848	18.0	2520	29.3936	0.9801	0.9760	0.9780	0.9929
8.9848	19.0	2660	31.2320	0.9796	0.9796	0.9796	0.9944
8.9848	20.0	2800	34.0474	0.9849	0.9802	0.9825	0.9943
8.9848	21.0	2940	32.9968	0.9849	0.9826	0.9838	0.9948
8.2401	22.0	3080	39.6873	0.9819	0.9808	0.9814	0.9946
8.2401	23.0	3220	42.7506	0.9819	0.9802	0.9811	0.9945
8.2401	24.0	3360	33.8886	0.9856	0.9862	0.9859	0.9954
7.099	25.0	3500	36.8275	0.9819	0.9808	0.9814	0.9941
7.099	26.0	3640	36.7838	0.9831	0.9814	0.9823	0.9951
7.099	27.0	3780	39.2226	0.9813	0.9790	0.9801	0.9947
7.099	28.0	3920	39.2492	0.9843	0.9820	0.9832	0.9949
5.6646	29.0	4060	41.4139	0.9790	0.9790	0.9790	0.9944
5.6646	30.0	4200	41.4583	0.9838	0.9826	0.9832	0.9949
5.6646	31.0	4340	47.1872	0.9801	0.9778	0.9789	0.9941
5.6646	32.0	4480	41.3073	0.9862	0.9844	0.9853	0.9956
5.304	33.0	4620	44.8882	0.9796	0.9790	0.9793	0.9945
5.304	34.0	4760	52.3203	0.9783	0.9772	0.9778	0.9941
5.304	35.0	4900	43.9140	0.9825	0.9808	0.9817	0.9951
4.7574	36.0	5040	46.8215	0.9819	0.9802	0.9811	0.9947
4.7574	37.0	5180	39.5738	0.9867	0.9844	0.9856	0.9959
4.7574	38.0	5320	39.9370	0.9837	0.9814	0.9826	0.9955
4.7574	39.0	5460	40.4614	0.9856	0.9844	0.9850	0.9956
3.8125	40.0	5600	38.6418	0.9885	0.9850	0.9868	0.9959
3.8125	41.0	5740	42.7438	0.9813	0.9796	0.9805	0.9947
3.8125	42.0	5880	52.7676	0.9689	0.9730	0.9709	0.9940
3.2902	43.0	6020	38.5737	0.9825	0.9808	0.9817	0.9953
3.2902	44.0	6160	42.4615	0.9868	0.9850	0.9859	0.9952
3.2902	45.0	6300	43.5099	0.9856	0.9838	0.9847	0.9956
3.2902	46.0	6440	45.0846	0.9837	0.9820	0.9829	0.9952
4.0467	47.0	6580	41.7571	0.9862	0.9850	0.9856	0.9955
4.0467	48.0	6720	50.8592	0.9807	0.9778	0.9792	0.9945
4.0467	49.0	6860	42.3155	0.9886	0.9862	0.9874	0.9956
2.3503	50.0	7000	45.7602	0.9873	0.9850	0.9862	0.9952
2.3503	51.0	7140	43.4314	0.9856	0.9838	0.9847	0.9953
2.3503	52.0	7280	47.4167	0.9813	0.9790	0.9801	0.9949
2.3503	53.0	7420	46.8868	0.9838	0.9826	0.9832	0.9952
2.841	54.0	7560	50.8428	0.9843	0.9814	0.9828	0.9950
2.841	55.0	7700	49.0097	0.9825	0.9808	0.9817	0.9949
2.841	56.0	7840	49.0165	0.9831	0.9802	0.9816	0.9950
2.841	57.0	7980	46.3213	0.9838	0.9826	0.9832	0.9953
1.8064	58.0	8120	49.3268	0.9825	0.9790	0.9807	0.9946
1.8064	59.0	8260	48.1988	0.9849	0.9814	0.9831	0.9952
1.8064	60.0	8400	46.5527	0.9838	0.9826	0.9832	0.9955
1.5941	61.0	8540	57.5747	0.9807	0.9790	0.9798	0.9942
1.5941	62.0	8680	56.6894	0.9801	0.9790	0.9796	0.9945
1.5941	63.0	8820	58.1243	0.9808	0.9802	0.9805	0.9945
1.5941	64.0	8960	53.2165	0.9837	0.9808	0.9822	0.9951
1.5057	65.0	9100	52.2484	0.9832	0.9820	0.9826	0.9949
1.5057	66.0	9240	49.2435	0.9837	0.9814	0.9826	0.9951
1.5057	67.0	9380	51.2186	0.9796	0.9790	0.9793	0.9948
1.5084	68.0	9520	54.1799	0.9825	0.9808	0.9817	0.9947
1.5084	69.0	9660	56.3696	0.9807	0.9778	0.9792	0.9945
1.5084	70.0	9800	52.6295	0.9837	0.9802	0.9819	0.9948
1.5084	71.0	9940	51.2577	0.9825	0.9790	0.9807	0.9950
1.0448	72.0	10080	56.0093	0.9807	0.9790	0.9798	0.9945
1.0448	73.0	10220	50.7540	0.9831	0.9808	0.9819	0.9951
1.0448	74.0	10360	52.9783	0.9819	0.9790	0.9804	0.9947

Framework versions

• Transformers 4.32.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3