t5-small-thaisum-title-mt5tokenizer

This model is a fine-tuned version of Nopphakorn/t5-small-thaisum-title-mt5tokenizer on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 3.4541
Rouge1: 0.0552
Rouge2: 0.0069
Rougel: 0.0547
Rougelsum: 0.0546
Gen Len: 18.9956

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: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
4.7436	1.0	765	4.8643	0.0029	0.0	0.0029	0.0029	18.9897
4.7069	2.0	1530	4.8482	0.0114	0.0015	0.0111	0.0115	18.9941
4.6588	3.0	2295	4.8272	0.0054	0.0	0.0055	0.0055	18.9883
4.6664	4.0	3060	4.8044	0.0116	0.0	0.0113	0.0116	18.9853
4.6283	5.0	3825	4.7903	0.0094	0.0	0.0095	0.0093	18.9868
4.6145	6.0	4590	4.7669	0.009	0.0	0.0088	0.0088	18.9941
4.5759	7.0	5355	4.7432	0.008	0.0	0.0079	0.0079	18.9883
4.5419	8.0	6120	4.7275	0.012	0.0015	0.012	0.012	18.9941
4.5486	9.0	6885	4.7043	0.0098	0.0015	0.0099	0.0097	18.9941
4.5095	10.0	7650	4.6773	0.0085	0.0	0.0088	0.0086	18.9941
4.4682	11.0	8415	4.6561	0.0115	0.0	0.0112	0.0114	18.9927
4.4783	12.0	9180	4.6326	0.0063	0.0	0.0063	0.0061	18.9868
4.4825	13.0	9945	4.6069	0.0111	0.0	0.0109	0.0111	18.9838
4.4455	14.0	10710	4.5836	0.0086	0.0	0.0084	0.0085	18.9941
4.4328	15.0	11475	4.5587	0.0089	0.0	0.0091	0.0088	18.9912
4.3982	16.0	12240	4.5294	0.0111	0.0	0.0112	0.0112	18.9868
4.3463	17.0	13005	4.5069	0.0135	0.0	0.0137	0.0137	18.9853
4.3379	18.0	13770	4.4717	0.011	0.0007	0.0109	0.011	18.9897
4.3303	19.0	14535	4.4460	0.0117	0.0007	0.0119	0.0118	18.9853
4.2983	20.0	15300	4.4110	0.0107	0.0	0.0109	0.0108	18.9868
4.265	21.0	16065	4.3800	0.0095	0.0	0.0096	0.0095	18.9941
4.2507	22.0	16830	4.3528	0.0105	0.0	0.0106	0.0104	18.9941
4.2184	23.0	17595	4.3246	0.0125	0.0007	0.0128	0.0125	18.9941
4.1746	24.0	18360	4.3013	0.0116	0.0	0.0118	0.0115	19.0
4.1744	25.0	19125	4.2724	0.0127	0.0	0.0128	0.0127	18.9956
4.1123	26.0	19890	4.2408	0.0127	0.0	0.0128	0.0128	19.0
4.131	27.0	20655	4.2045	0.0118	0.0	0.0121	0.0118	18.9897
4.0725	28.0	21420	4.1815	0.0112	0.0	0.0113	0.011	19.0
4.0439	29.0	22185	4.1452	0.0154	0.0024	0.0154	0.0152	19.0
4.0814	30.0	22950	4.1151	0.0142	0.0007	0.0143	0.014	19.0
4.009	31.0	23715	4.0963	0.015	0.0007	0.015	0.015	18.9985
4.0326	32.0	24480	4.0679	0.016	0.0	0.016	0.0161	19.0
3.948	33.0	25245	4.0460	0.0137	0.0015	0.0133	0.0132	18.9971
3.9401	34.0	26010	4.0070	0.0162	0.0024	0.0161	0.016	18.9985
3.9111	35.0	26775	3.9874	0.0209	0.0042	0.0209	0.021	18.9985
3.899	36.0	27540	3.9650	0.0185	0.0	0.0186	0.0186	18.9985
3.8821	37.0	28305	3.9457	0.0236	0.0029	0.024	0.024	19.0
3.8687	38.0	29070	3.9105	0.0241	0.0034	0.0239	0.0238	19.0
3.8076	39.0	29835	3.9029	0.0199	0.002	0.0199	0.0201	19.0
3.8063	40.0	30600	3.8750	0.0251	0.0034	0.0244	0.0246	19.0
3.7896	41.0	31365	3.8448	0.025	0.0028	0.0248	0.0249	18.9941
3.7672	42.0	32130	3.8287	0.0336	0.0029	0.0333	0.0332	19.0
3.7919	43.0	32895	3.8022	0.0268	0.0039	0.0268	0.0268	18.9956
3.75	44.0	33660	3.7723	0.0286	0.0044	0.0286	0.0284	18.9971
3.7263	45.0	34425	3.7630	0.0308	0.0039	0.0308	0.0307	19.0
3.7053	46.0	35190	3.7412	0.0341	0.0037	0.0335	0.0335	19.0
3.7022	47.0	35955	3.7214	0.0347	0.0044	0.0335	0.0336	18.9897
3.6528	48.0	36720	3.7059	0.0318	0.0044	0.032	0.032	19.0
3.6614	49.0	37485	3.6833	0.0313	0.0044	0.031	0.0309	18.9956
3.6339	50.0	38250	3.6691	0.0357	0.0051	0.0353	0.0352	18.9853
3.6153	51.0	39015	3.6500	0.0373	0.0044	0.0363	0.0365	18.9912
3.6083	52.0	39780	3.6360	0.0358	0.0051	0.0354	0.0354	18.9985
3.5857	53.0	40545	3.6272	0.0409	0.0044	0.0397	0.0396	19.0
3.5903	54.0	41310	3.6141	0.0455	0.0039	0.044	0.0439	18.9956
3.5429	55.0	42075	3.6044	0.0405	0.0054	0.0393	0.0394	18.9883
3.5526	56.0	42840	3.5933	0.0379	0.0049	0.037	0.0367	18.9883
3.5075	57.0	43605	3.5820	0.0431	0.0071	0.041	0.041	18.9985
3.5233	58.0	44370	3.5698	0.045	0.0064	0.0433	0.0434	18.9897
3.5022	59.0	45135	3.5680	0.0432	0.0061	0.0419	0.0416	18.9941
3.5258	60.0	45900	3.5604	0.047	0.0069	0.0452	0.0453	18.9956
3.4763	61.0	46665	3.5532	0.0465	0.0069	0.0452	0.0451	18.9985
3.4591	62.0	47430	3.5468	0.0429	0.0078	0.0425	0.0422	18.9971
3.471	63.0	48195	3.5359	0.0426	0.0078	0.0428	0.0426	18.9971
3.4671	64.0	48960	3.5348	0.0441	0.0064	0.0437	0.0436	18.9941
3.4588	65.0	49725	3.5291	0.0436	0.0064	0.0427	0.0426	18.9941
3.4214	66.0	50490	3.5168	0.0409	0.0071	0.0407	0.0408	18.9956
3.4531	67.0	51255	3.5091	0.0476	0.0082	0.0477	0.0479	18.9956
3.3936	68.0	52020	3.5016	0.044	0.0073	0.0441	0.0442	18.9956
3.4113	69.0	52785	3.5028	0.0473	0.0069	0.0472	0.0472	18.9956
3.4092	70.0	53550	3.4993	0.0483	0.0078	0.0488	0.0485	18.9985
3.3847	71.0	54315	3.4959	0.053	0.0078	0.0524	0.0524	18.9956
3.4099	72.0	55080	3.4906	0.0549	0.0069	0.0541	0.0542	18.9985
3.3774	73.0	55845	3.4821	0.0527	0.0064	0.052	0.052	18.9971
3.3677	74.0	56610	3.4790	0.0542	0.0069	0.0534	0.0534	18.9956
3.3707	75.0	57375	3.4747	0.0562	0.0069	0.0556	0.0557	18.9956
3.3953	76.0	58140	3.4713	0.0567	0.0069	0.056	0.056	18.9956
3.3767	77.0	58905	3.4695	0.0559	0.0069	0.0552	0.0551	18.9956
3.3455	78.0	59670	3.4668	0.0518	0.0069	0.0514	0.0508	18.9985
3.3749	79.0	60435	3.4648	0.055	0.0069	0.0546	0.0545	18.9941
3.3447	80.0	61200	3.4648	0.0534	0.0069	0.0529	0.0526	18.9956
3.3892	81.0	61965	3.4643	0.0572	0.0078	0.0568	0.0564	18.9985
3.3681	82.0	62730	3.4640	0.0545	0.0069	0.0542	0.0537	18.9956
3.3186	83.0	63495	3.4595	0.0574	0.0069	0.0569	0.0569	18.9956
3.3422	84.0	64260	3.4598	0.0553	0.0069	0.0549	0.0548	18.9956
3.3511	85.0	65025	3.4565	0.0563	0.0078	0.0561	0.0561	18.9956
3.3469	86.0	65790	3.4576	0.0569	0.0069	0.0562	0.0563	18.9956
3.345	87.0	66555	3.4579	0.0553	0.0069	0.0549	0.0548	18.9956
3.3611	88.0	67320	3.4558	0.0553	0.0069	0.0549	0.0548	18.9956
3.3423	89.0	68085	3.4559	0.0569	0.0069	0.0562	0.0563	18.9956
3.3575	90.0	68850	3.4560	0.0553	0.0069	0.0549	0.0548	18.9956
3.3322	91.0	69615	3.4560	0.0569	0.0069	0.0562	0.0563	18.9956
3.3303	92.0	70380	3.4551	0.0569	0.0069	0.0562	0.0563	18.9956
3.3676	93.0	71145	3.4542	0.0569	0.0069	0.0562	0.0563	18.9956
3.3219	94.0	71910	3.4541	0.0552	0.0069	0.0547	0.0546	18.9956
3.3563	95.0	72675	3.4540	0.0569	0.0069	0.0562	0.0563	18.9956
3.3616	96.0	73440	3.4541	0.0569	0.0069	0.0562	0.0563	18.9956
3.3417	97.0	74205	3.4543	0.0552	0.0069	0.0547	0.0546	18.9956
3.3683	98.0	74970	3.4541	0.0552	0.0069	0.0547	0.0546	18.9956
3.3402	99.0	75735	3.4541	0.0552	0.0069	0.0547	0.0546	18.9956
3.3413	100.0	76500	3.4541	0.0552	0.0069	0.0547	0.0546	18.9956

Framework versions

Transformers 4.29.2
Pytorch 2.0.0+cu118
Datasets 2.12.0
Tokenizers 0.13.3