mt5-small-thaisum

Nopphakorn/mt5-small-thaisum

mt5-small-thaisum is a machine learning model from Nopphakorn. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

3% of all-time downloads

All-time downloads

339

Public

Repo size

46.8 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin1.2 GB · 100%

At a glance

Library: transformers
License: apache-2.0
Model type: mt5
Access: Public
Created: May 14, 2023
Updated: May 14, 2023
SHA: 4cf7bff4

Model card

From the Hugging Face model README

mt5-small-thaisum

This model is a fine-tuned version of google/mt5-small on the None dataset. It achieves the following results on the evaluation set:

Loss: 3.0602
Rouge1: 0.0804
Rouge2: 0.0092
Rougel: 0.079
Rougelsum: 0.0804
Gen Len: 19.0

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: 8
eval_batch_size: 8
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	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	200	2.9811	0.0876	0.0172	0.0883	0.0879	18.3775
No log	2.0	400	2.8529	0.0814	0.0205	0.0824	0.0818	18.6075
4.6025	3.0	600	2.7693	0.08	0.0182	0.0801	0.0796	18.7575
4.6025	4.0	800	2.6774	0.0874	0.0195	0.0859	0.0861	18.8075
3.1265	5.0	1000	2.6566	0.0924	0.0243	0.0903	0.0906	18.92
3.1265	6.0	1200	2.6393	0.0993	0.0232	0.0994	0.0994	18.915
3.1265	7.0	1400	2.6257	0.092	0.0191	0.0914	0.0915	18.9725
2.8328	8.0	1600	2.6159	0.095	0.0187	0.0939	0.0944	18.9575
2.8328	9.0	1800	2.5708	0.0978	0.0223	0.0975	0.0976	18.9575
2.6385	10.0	2000	2.5637	0.09	0.0188	0.0902	0.0907	18.965
2.6385	11.0	2200	2.5621	0.0989	0.0179	0.0974	0.0979	18.975
2.6385	12.0	2400	2.5490	0.1002	0.0206	0.0996	0.0999	18.98
2.4812	13.0	2600	2.5260	0.1001	0.0175	0.0987	0.0994	18.9875
2.4812	14.0	2800	2.5626	0.0987	0.0234	0.098	0.0986	18.9875
2.3701	15.0	3000	2.5042	0.0965	0.021	0.0955	0.0965	18.99
2.3701	16.0	3200	2.5017	0.1038	0.021	0.1024	0.1034	18.99
2.3701	17.0	3400	2.5204	0.0914	0.0192	0.0905	0.0909	18.99
2.2397	18.0	3600	2.5311	0.0909	0.01	0.0896	0.0901	18.99
2.2397	19.0	3800	2.5149	0.0938	0.0131	0.0932	0.0939	18.99
2.1529	20.0	4000	2.5644	0.0944	0.016	0.0934	0.0949	18.99
2.1529	21.0	4200	2.5408	0.0918	0.0138	0.0906	0.0905	18.99
2.1529	22.0	4400	2.5415	0.083	0.0096	0.0823	0.0829	18.99
2.0546	23.0	4600	2.5564	0.0927	0.0154	0.092	0.0921	18.99
2.0546	24.0	4800	2.5778	0.0877	0.015	0.0868	0.0874	18.9875
1.9784	25.0	5000	2.5437	0.09	0.0144	0.0893	0.0902	18.99
1.9784	26.0	5200	2.5665	0.0838	0.0096	0.0826	0.0831	18.985
1.9784	27.0	5400	2.5867	0.0922	0.0129	0.0909	0.0921	18.99
1.8855	28.0	5600	2.5676	0.0887	0.0167	0.0882	0.0885	18.995
1.8855	29.0	5800	2.5806	0.0889	0.0162	0.0876	0.0881	18.9975
1.8416	30.0	6000	2.5722	0.0951	0.0183	0.0936	0.095	19.0
1.8416	31.0	6200	2.5865	0.0866	0.0119	0.0857	0.0862	18.99
1.8416	32.0	6400	2.5779	0.0956	0.0204	0.0933	0.0939	19.0
1.767	33.0	6600	2.6038	0.0954	0.02	0.0942	0.0962	18.99
1.767	34.0	6800	2.6091	0.0951	0.0192	0.0941	0.0953	19.0
1.6979	35.0	7000	2.6599	0.0813	0.0138	0.0802	0.0805	18.9875
1.6979	36.0	7200	2.6338	0.0905	0.015	0.0898	0.0901	18.99
1.6979	37.0	7400	2.6515	0.0898	0.015	0.0892	0.0894	18.985
1.6445	38.0	7600	2.6514	0.0797	0.0142	0.0788	0.079	18.985
1.6445	39.0	7800	2.6418	0.0905	0.0196	0.0884	0.089	19.0
1.5932	40.0	8000	2.6520	0.0764	0.0146	0.0758	0.0766	19.0
1.5932	41.0	8200	2.6709	0.0793	0.01	0.0773	0.0788	18.99
1.5932	42.0	8400	2.6838	0.0838	0.0162	0.0824	0.0836	18.99
1.5379	43.0	8600	2.6731	0.0876	0.0154	0.086	0.087	18.9875
1.5379	44.0	8800	2.6804	0.0902	0.0161	0.0889	0.0898	18.99
1.4868	45.0	9000	2.6993	0.0852	0.0158	0.0839	0.0849	18.99
1.4868	46.0	9200	2.6954	0.0785	0.0117	0.0778	0.0778	19.0
1.4868	47.0	9400	2.7108	0.0837	0.015	0.0823	0.0828	18.9875
1.4401	48.0	9600	2.7319	0.09	0.0154	0.0886	0.0888	18.9975
1.4401	49.0	9800	2.7670	0.0924	0.0175	0.0912	0.0919	18.99
1.4115	50.0	10000	2.7521	0.0853	0.017	0.0838	0.0846	19.0
1.4115	51.0	10200	2.7725	0.0885	0.0198	0.0869	0.0879	18.99
1.4115	52.0	10400	2.7646	0.0825	0.0158	0.0815	0.0824	19.0
1.3594	53.0	10600	2.7719	0.0842	0.0112	0.083	0.0835	18.99
1.3594	54.0	10800	2.8063	0.0884	0.0158	0.0868	0.0874	18.99
1.3279	55.0	11000	2.8011	0.0926	0.0175	0.0906	0.091	19.0
1.3279	56.0	11200	2.8125	0.0897	0.0125	0.0879	0.0892	19.0
1.3279	57.0	11400	2.8529	0.088	0.0158	0.0861	0.0863	19.0
1.2902	58.0	11600	2.8338	0.0952	0.0133	0.093	0.0939	19.0
1.2902	59.0	11800	2.8260	0.091	0.0167	0.0885	0.0894	19.0
1.2662	60.0	12000	2.8383	0.0948	0.014	0.0916	0.0925	19.0
1.2662	61.0	12200	2.8589	0.0861	0.0158	0.0835	0.0848	19.0
1.2662	62.0	12400	2.8772	0.0855	0.0158	0.0831	0.084	18.99
1.2309	63.0	12600	2.8566	0.0847	0.0158	0.0836	0.0833	19.0
1.2309	64.0	12800	2.8800	0.0848	0.0158	0.0833	0.0841	19.0
1.1995	65.0	13000	2.8895	0.0855	0.0112	0.0841	0.0852	19.0
1.1995	66.0	13200	2.8857	0.0865	0.0133	0.0845	0.0856	19.0
1.1995	67.0	13400	2.8999	0.0891	0.015	0.0875	0.0885	19.0
1.1799	68.0	13600	2.9295	0.0866	0.0133	0.0838	0.0856	18.9875
1.1799	69.0	13800	2.9197	0.0782	0.0083	0.0771	0.0781	18.9875
1.1527	70.0	14000	2.9388	0.0789	0.0125	0.0778	0.0784	18.9975
1.1527	71.0	14200	2.9264	0.0736	0.0075	0.0725	0.0736	19.0
1.1527	72.0	14400	2.9597	0.0862	0.01	0.0848	0.0859	19.0
1.1322	73.0	14600	2.9623	0.084	0.0092	0.083	0.0839	19.0
1.1322	74.0	14800	2.9681	0.08	0.011	0.0792	0.0798	19.0
1.1075	75.0	15000	2.9713	0.0844	0.0142	0.0825	0.0827	19.0
1.1075	76.0	15200	2.9669	0.0815	0.0133	0.0803	0.0808	19.0
1.1075	77.0	15400	2.9680	0.0718	0.0092	0.0712	0.0723	18.99
1.0893	78.0	15600	2.9932	0.0811	0.0108	0.0803	0.0812	18.99
1.0893	79.0	15800	2.9979	0.0832	0.0133	0.082	0.083	18.99
1.0733	80.0	16000	2.9997	0.0863	0.0142	0.085	0.0858	18.99
1.0733	81.0	16200	2.9965	0.086	0.0162	0.0845	0.0858	18.99
1.0733	82.0	16400	3.0170	0.0813	0.0125	0.0788	0.0799	18.99
1.0617	83.0	16600	2.9955	0.0865	0.0125	0.0839	0.0851	18.99
1.0617	84.0	16800	3.0299	0.0912	0.015	0.0903	0.0912	19.0
1.0417	85.0	17000	3.0333	0.089	0.0142	0.0877	0.0878	19.0
1.0417	86.0	17200	3.0282	0.0948	0.0158	0.0941	0.0938	19.0
1.0417	87.0	17400	3.0416	0.0906	0.015	0.0897	0.0908	19.0
1.0316	88.0	17600	3.0376	0.0874	0.0142	0.0871	0.0876	19.0
1.0316	89.0	17800	3.0287	0.0873	0.0142	0.0864	0.0871	19.0
1.0285	90.0	18000	3.0384	0.0798	0.0142	0.079	0.0797	19.0
1.0285	91.0	18200	3.0465	0.0759	0.01	0.0751	0.076	19.0
1.0285	92.0	18400	3.0444	0.0797	0.0117	0.0788	0.0794	19.0
1.0218	93.0	18600	3.0524	0.0751	0.0075	0.0744	0.0754	19.0
1.0218	94.0	18800	3.0524	0.0744	0.0075	0.0736	0.0746	19.0
1.0086	95.0	19000	3.0466	0.0815	0.0092	0.079	0.0803	19.0
1.0086	96.0	19200	3.0561	0.0804	0.0092	0.079	0.0804	19.0
1.0086	97.0	19400	3.0596	0.0774	0.0092	0.0761	0.0774	19.0
1.0027	98.0	19600	3.0548	0.0792	0.0092	0.0778	0.0789	19.0
1.0027	99.0	19800	3.0601	0.0804	0.0092	0.079	0.0804	19.0
1.0023	100.0	20000	3.0602	0.0804	0.0092	0.079	0.0804	19.0

Framework versions

Transformers 4.29.1
Pytorch 2.0.0+cu118
Datasets 2.12.0
Tokenizers 0.13.3

Library: transformers
Type: mt5
License: apache-2.0
Created: May 14, 2023
Updated: May 14, 2023