ner_roberta

This model is a fine-tuned version of FacebookAI/roberta-base on the hts98/UIT dataset. It achieves the following results on the evaluation set:

Loss: 1.7625
Precision: 0.6042
Recall: 0.6798
F1: 0.6398
Accuracy: 0.8047

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: 3e-05
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 120.0

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	122	1.1792	0.2153	0.3176	0.2566	0.6467
No log	2.0	244	0.9056	0.3515	0.4828	0.4069	0.7298
No log	3.0	366	0.8212	0.4012	0.5341	0.4582	0.7447
No log	4.0	488	0.7602	0.4275	0.5653	0.4868	0.7638
1.0934	5.0	610	0.7661	0.4252	0.5873	0.4933	0.7646
1.0934	6.0	732	0.7474	0.4789	0.6080	0.5358	0.7731
1.0934	7.0	854	0.7387	0.4920	0.5979	0.5398	0.7785
1.0934	8.0	976	0.7482	0.4917	0.6168	0.5472	0.7814
0.5404	9.0	1098	0.7774	0.4963	0.6266	0.5539	0.7782
0.5404	10.0	1220	0.7820	0.5074	0.6246	0.5599	0.7802
0.5404	11.0	1342	0.7770	0.5092	0.6363	0.5657	0.7817
0.5404	12.0	1464	0.8045	0.5340	0.6303	0.5781	0.7849
0.3509	13.0	1586	0.8088	0.5344	0.6478	0.5856	0.7871
0.3509	14.0	1708	0.8470	0.5049	0.6432	0.5657	0.7768
0.3509	15.0	1830	0.8358	0.5298	0.6415	0.5803	0.7846
0.3509	16.0	1952	0.8826	0.5216	0.6369	0.5735	0.7889
0.2458	17.0	2074	0.8950	0.5477	0.6400	0.5903	0.7904
0.2458	18.0	2196	0.8846	0.5212	0.6506	0.5788	0.7900
0.2458	19.0	2318	0.8888	0.5335	0.6455	0.5842	0.7904
0.2458	20.0	2440	0.8934	0.5345	0.6346	0.5803	0.7915
0.1765	21.0	2562	0.9482	0.5459	0.6449	0.5913	0.7933
0.1765	22.0	2684	0.9499	0.5462	0.6481	0.5928	0.7959
0.1765	23.0	2806	0.9826	0.5495	0.6369	0.5900	0.7869
0.1765	24.0	2928	0.9815	0.5714	0.6475	0.6071	0.7973
0.1273	25.0	3050	1.0080	0.5618	0.6498	0.6026	0.7971
0.1273	26.0	3172	1.0463	0.5472	0.6544	0.5960	0.7940
0.1273	27.0	3294	1.0349	0.5574	0.6475	0.5991	0.7966
0.1273	28.0	3416	1.0559	0.5496	0.6523	0.5966	0.7898
0.0951	29.0	3538	1.0901	0.5433	0.6523	0.5928	0.7918
0.0951	30.0	3660	1.1400	0.5297	0.6532	0.5850	0.7841
0.0951	31.0	3782	1.1601	0.5624	0.6337	0.5959	0.7916
0.0951	32.0	3904	1.1359	0.5498	0.6532	0.5970	0.7883
0.0717	33.0	4026	1.1269	0.5625	0.6564	0.6058	0.7965
0.0717	34.0	4148	1.1758	0.5679	0.6443	0.6037	0.7967
0.0717	35.0	4270	1.1870	0.5493	0.6506	0.5957	0.7898
0.0717	36.0	4392	1.1296	0.5509	0.6558	0.5988	0.7928
0.0552	37.0	4514	1.2164	0.5415	0.6564	0.5934	0.7903
0.0552	38.0	4636	1.2047	0.5516	0.6581	0.6002	0.7943
0.0552	39.0	4758	1.2364	0.5641	0.6604	0.6084	0.7949
0.0552	40.0	4880	1.2481	0.5573	0.6598	0.6042	0.7945
0.0432	41.0	5002	1.2768	0.5684	0.6452	0.6043	0.7926
0.0432	42.0	5124	1.2605	0.5639	0.6595	0.6080	0.7958
0.0432	43.0	5246	1.2495	0.5710	0.6607	0.6126	0.7975
0.0432	44.0	5368	1.2718	0.5762	0.6486	0.6103	0.7955
0.0432	45.0	5490	1.2998	0.5725	0.6512	0.6093	0.8003
0.0331	46.0	5612	1.3469	0.5620	0.6455	0.6008	0.7946
0.0331	47.0	5734	1.3357	0.5722	0.6604	0.6131	0.8010
0.0331	48.0	5856	1.3576	0.5583	0.6578	0.6040	0.7936
0.0331	49.0	5978	1.3397	0.5766	0.6584	0.6148	0.7985
0.0265	50.0	6100	1.3641	0.5671	0.6549	0.6078	0.7961
0.0265	51.0	6222	1.3727	0.5637	0.6546	0.6058	0.7938
0.0265	52.0	6344	1.4025	0.5624	0.6621	0.6082	0.7927
0.0265	53.0	6466	1.3991	0.5672	0.6561	0.6084	0.7916
0.0212	54.0	6588	1.4268	0.5664	0.6655	0.6120	0.7953
0.0212	55.0	6710	1.4377	0.5636	0.6584	0.6073	0.7944
0.0212	56.0	6832	1.4307	0.5689	0.6607	0.6114	0.7953
0.0212	57.0	6954	1.4773	0.5678	0.6472	0.6049	0.7917
0.0171	58.0	7076	1.4626	0.5928	0.6532	0.6215	0.7976
0.0171	59.0	7198	1.4489	0.5726	0.6569	0.6119	0.7991
0.0171	60.0	7320	1.4479	0.5834	0.6592	0.6190	0.8010
0.0171	61.0	7442	1.4649	0.5828	0.6523	0.6156	0.7976
0.0142	62.0	7564	1.5170	0.5726	0.6698	0.6174	0.8006
0.0142	63.0	7686	1.4866	0.5776	0.6661	0.6187	0.7985
0.0142	64.0	7808	1.5446	0.5788	0.6604	0.6169	0.8010
0.0142	65.0	7930	1.5566	0.5687	0.6604	0.6111	0.7935
0.0114	66.0	8052	1.5454	0.5896	0.6632	0.6243	0.7959
0.0114	67.0	8174	1.5341	0.6015	0.6670	0.6325	0.7998
0.0114	68.0	8296	1.5298	0.5864	0.6569	0.6197	0.7963
0.0114	69.0	8418	1.5694	0.5773	0.6638	0.6176	0.7944
0.0101	70.0	8540	1.5914	0.5805	0.6647	0.6198	0.7977
0.0101	71.0	8662	1.5686	0.5728	0.6592	0.6130	0.7957
0.0101	72.0	8784	1.6199	0.5647	0.6695	0.6127	0.7949
0.0101	73.0	8906	1.6344	0.5848	0.6667	0.6230	0.7944
0.0079	74.0	9028	1.5580	0.5915	0.6721	0.6292	0.7970
0.0079	75.0	9150	1.6272	0.6054	0.6655	0.6340	0.8006
0.0079	76.0	9272	1.6267	0.5795	0.6629	0.6184	0.7968
0.0079	77.0	9394	1.6501	0.5758	0.6704	0.6195	0.7959
0.0065	78.0	9516	1.6222	0.5959	0.6690	0.6303	0.7995
0.0065	79.0	9638	1.6543	0.5878	0.6690	0.6258	0.7966
0.0065	80.0	9760	1.6054	0.5922	0.6675	0.6276	0.8009
0.0065	81.0	9882	1.6387	0.5941	0.6690	0.6293	0.8008
0.0053	82.0	10004	1.6453	0.6098	0.6712	0.6390	0.8047
0.0053	83.0	10126	1.6794	0.5803	0.6684	0.6212	0.8005
0.0053	84.0	10248	1.7006	0.5979	0.6690	0.6314	0.7990
0.0053	85.0	10370	1.6820	0.5928	0.6715	0.6297	0.7989
0.0053	86.0	10492	1.6995	0.5920	0.6698	0.6285	0.7983
0.0045	87.0	10614	1.6652	0.5923	0.6624	0.6254	0.8005
0.0045	88.0	10736	1.7196	0.5919	0.6658	0.6267	0.7991
0.0045	89.0	10858	1.6730	0.5953	0.6629	0.6273	0.8054
0.0045	90.0	10980	1.7092	0.5966	0.6747	0.6332	0.8023
0.0037	91.0	11102	1.7260	0.6035	0.6678	0.6340	0.8010
0.0037	92.0	11224	1.7106	0.5998	0.6670	0.6316	0.8030
0.0037	93.0	11346	1.7096	0.6047	0.6747	0.6378	0.8028
0.0037	94.0	11468	1.7220	0.5986	0.6770	0.6354	0.8010
0.0032	95.0	11590	1.7394	0.5966	0.6790	0.6351	0.7994
0.0032	96.0	11712	1.7257	0.6074	0.6744	0.6392	0.8005
0.0032	97.0	11834	1.7008	0.6046	0.6687	0.6350	0.8039
0.0032	98.0	11956	1.7482	0.6029	0.6718	0.6355	0.8032
0.0028	99.0	12078	1.7570	0.5988	0.6690	0.6319	0.8030
0.0028	100.0	12200	1.7332	0.5980	0.6735	0.6335	0.8026
0.0028	101.0	12322	1.7491	0.5902	0.6710	0.6280	0.8011
0.0028	102.0	12444	1.7542	0.6003	0.6735	0.6348	0.8033
0.0021	103.0	12566	1.7371	0.5956	0.6698	0.6305	0.8040
0.0021	104.0	12688	1.7719	0.5914	0.6678	0.6273	0.8007
0.0021	105.0	12810	1.7473	0.5982	0.6667	0.6306	0.8045
0.0021	106.0	12932	1.7518	0.6002	0.6767	0.6362	0.8040
0.0019	107.0	13054	1.7628	0.6010	0.6750	0.6358	0.8048
0.0019	108.0	13176	1.8080	0.5970	0.6770	0.6345	0.7965
0.0019	109.0	13298	1.8028	0.5961	0.6767	0.6339	0.7986
0.0019	110.0	13420	1.7820	0.5996	0.6733	0.6343	0.8030
0.0015	111.0	13542	1.7890	0.6024	0.6764	0.6373	0.8030
0.0015	112.0	13664	1.7686	0.6071	0.6761	0.6398	0.8040
0.0015	113.0	13786	1.7625	0.6042	0.6798	0.6398	0.8047
0.0015	114.0	13908	1.7637	0.6054	0.6735	0.6377	0.8039
0.0013	115.0	14030	1.7680	0.6041	0.6770	0.6385	0.8038
0.0013	116.0	14152	1.7831	0.6001	0.6781	0.6367	0.8029
0.0013	117.0	14274	1.7854	0.5994	0.6758	0.6353	0.8021
0.0013	118.0	14396	1.7762	0.6011	0.6741	0.6355	0.8034
0.0012	119.0	14518	1.7778	0.6011	0.6750	0.6359	0.8028
0.0012	120.0	14640	1.7774	0.6016	0.6747	0.6361	0.8029

Framework versions

Transformers 4.32.0.dev0
Pytorch 2.1.0+cu121
Datasets 3.1.0
Tokenizers 0.13.3

hts98/ner_roberta

ner_roberta

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种