roberta-base-finetuned-ner-cadec-no-iob

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

Loss: 0.4142
Precision: 0.6452
Recall: 0.6860
F1: 0.6650
Accuracy: 0.9380
Adr Precision: 0.5911
Adr Recall: 0.6557
Adr F1: 0.6217
Disease Precision: 0.4138
Disease Recall: 0.375
Disease F1: 0.3934
Drug Precision: 0.8962
Drug Recall: 0.9111
Drug F1: 0.9036
Finding Precision: 0.375
Finding Recall: 0.375
Finding F1: 0.375
Symptom Precision: 0.5833
Symptom Recall: 0.4828
Symptom F1: 0.5283
Macro Avg F1: 0.5644
Weighted Avg F1: 0.6650

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: 2e-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: linear
num_epochs: 35

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy	Adr Precision	Adr Recall	Adr F1	Disease Precision	Disease Recall	Disease F1	Drug Precision	Drug Recall	Drug F1	Finding Precision	Finding Recall	Finding F1	Symptom Precision	Symptom Recall	Symptom F1	Macro Avg F1	Weighted Avg F1
No log	1.0	125	0.2142	0.5325	0.6055	0.5667	0.9194	0.4548	0.5918	0.5143	0.4186	0.5625	0.48	0.8398	0.8444	0.8421	0.2857	0.0625	0.1026	0.0	0.0	0.0	0.3878	0.5537
No log	2.0	250	0.1798	0.6083	0.6557	0.6311	0.9339	0.5276	0.6309	0.5746	0.6842	0.4062	0.5098	0.8950	0.9	0.8975	0.32	0.25	0.2807	0.6667	0.2759	0.3902	0.5306	0.6291
No log	3.0	375	0.1910	0.5748	0.6029	0.5885	0.9282	0.5191	0.5320	0.5255	0.6	0.375	0.4615	0.8820	0.8722	0.8771	0.2927	0.375	0.3288	0.3051	0.6207	0.4091	0.5204	0.5935
0.1902	4.0	500	0.2013	0.5995	0.6398	0.6190	0.9311	0.5460	0.6	0.5717	0.25	0.0938	0.1364	0.8840	0.8889	0.8864	0.2632	0.4688	0.3371	0.6154	0.5517	0.5818	0.5027	0.6185
0.1902	5.0	625	0.2113	0.6161	0.6649	0.6396	0.9335	0.5515	0.6289	0.5877	0.5556	0.4688	0.5085	0.8852	0.9	0.8926	0.2857	0.25	0.2667	0.5185	0.4828	0.5	0.5511	0.6398
0.1902	6.0	750	0.1955	0.6223	0.6544	0.6379	0.9341	0.5541	0.6021	0.5771	0.5833	0.4375	0.5	0.8956	0.9056	0.9006	0.2857	0.375	0.3243	0.6818	0.5172	0.5882	0.5780	0.6404
0.1902	7.0	875	0.2226	0.6252	0.6491	0.6369	0.9343	0.5671	0.6186	0.5917	0.5556	0.1562	0.2439	0.8983	0.8833	0.8908	0.3061	0.4688	0.3704	0.5652	0.4483	0.5000	0.5193	0.6352
0.0648	8.0	1000	0.2345	0.6229	0.6755	0.6481	0.9363	0.5773	0.6392	0.6067	0.4138	0.375	0.3934	0.875	0.8944	0.8846	0.2973	0.3438	0.3188	0.5143	0.6207	0.5625	0.5532	0.6498
0.0648	9.0	1125	0.2316	0.6322	0.6689	0.65	0.9368	0.5851	0.6309	0.6071	0.5	0.3125	0.3846	0.8811	0.9056	0.8932	0.2766	0.4062	0.3291	0.5556	0.5172	0.5357	0.5499	0.6512
0.0648	10.0	1250	0.2944	0.6204	0.6491	0.6344	0.9320	0.5551	0.6021	0.5776	0.5789	0.3438	0.4314	0.8913	0.9111	0.9011	0.2619	0.3438	0.2973	0.6364	0.4828	0.5490	0.5513	0.6353
0.0648	11.0	1375	0.2660	0.6280	0.6794	0.6527	0.9353	0.5786	0.6454	0.6101	0.3824	0.4062	0.3939	0.8956	0.9056	0.9006	0.2812	0.2812	0.2812	0.5484	0.5862	0.5667	0.5505	0.6544
0.0284	12.0	1500	0.2819	0.6366	0.6702	0.6530	0.9355	0.5827	0.6392	0.6096	0.5	0.1875	0.2727	0.8956	0.9056	0.9006	0.3478	0.5	0.4103	0.5	0.4483	0.4727	0.5332	0.6508
0.0284	13.0	1625	0.3000	0.6326	0.6702	0.6509	0.9363	0.5736	0.6351	0.6027	0.4444	0.375	0.4068	0.8950	0.9	0.8975	0.3421	0.4062	0.3714	0.65	0.4483	0.5306	0.5618	0.6519
0.0284	14.0	1750	0.2996	0.6228	0.6491	0.6357	0.9363	0.5645	0.6041	0.5837	0.4783	0.3438	0.4	0.8743	0.8889	0.8815	0.2973	0.3438	0.3188	0.6071	0.5862	0.5965	0.5561	0.6360
0.0284	15.0	1875	0.3246	0.6311	0.6636	0.6469	0.9352	0.5788	0.6206	0.5990	0.45	0.2812	0.3462	0.9056	0.9056	0.9056	0.2683	0.3438	0.3014	0.5278	0.6552	0.5846	0.5473	0.6480
0.0136	16.0	2000	0.3305	0.6461	0.6623	0.6541	0.9377	0.5869	0.6268	0.6062	0.4545	0.3125	0.3704	0.9011	0.9111	0.9061	0.3448	0.3125	0.3279	0.5385	0.4828	0.5091	0.5439	0.6520
0.0136	17.0	2125	0.3181	0.6291	0.6781	0.6527	0.9375	0.5780	0.6495	0.6117	0.4231	0.3438	0.3793	0.9066	0.9167	0.9116	0.2857	0.3125	0.2985	0.4483	0.4483	0.4483	0.5299	0.6536
0.0136	18.0	2250	0.3414	0.6298	0.6755	0.6518	0.9362	0.5765	0.6371	0.6053	0.375	0.375	0.375	0.8962	0.9111	0.9036	0.3235	0.3438	0.3333	0.5714	0.5517	0.5614	0.5557	0.6532
0.0136	19.0	2375	0.3457	0.6302	0.6768	0.6527	0.9372	0.5877	0.6495	0.6170	0.3636	0.25	0.2963	0.8907	0.9056	0.8981	0.26	0.4062	0.3171	0.6087	0.4828	0.5385	0.5334	0.6546
0.0078	20.0	2500	0.3700	0.6367	0.6636	0.6499	0.9367	0.5805	0.6247	0.6018	0.3714	0.4062	0.3881	0.9016	0.9167	0.9091	0.3077	0.25	0.2759	0.5833	0.4828	0.5283	0.5406	0.6492
0.0078	21.0	2625	0.3772	0.6276	0.6715	0.6488	0.9325	0.5766	0.6289	0.6016	0.44	0.3438	0.3860	0.8919	0.9167	0.9041	0.2927	0.375	0.3288	0.5161	0.5517	0.5333	0.5508	0.6502
0.0078	22.0	2750	0.3622	0.6389	0.6768	0.6573	0.9345	0.5855	0.6495	0.6158	0.4333	0.4062	0.4194	0.8840	0.8889	0.8864	0.3333	0.3125	0.3226	0.625	0.5172	0.5660	0.5620	0.6575
0.0078	23.0	2875	0.3811	0.6304	0.6728	0.6509	0.9352	0.5765	0.6371	0.6053	0.4	0.375	0.3871	0.8804	0.9	0.8901	0.3438	0.3438	0.3438	0.5926	0.5517	0.5714	0.5595	0.6514
0.005	24.0	3000	0.3824	0.6322	0.6689	0.65	0.9353	0.5757	0.6351	0.6039	0.4286	0.375	0.4000	0.8901	0.9	0.8950	0.3226	0.3125	0.3175	0.5769	0.5172	0.5455	0.5524	0.6501
0.005	25.0	3125	0.3821	0.6297	0.6821	0.6548	0.9375	0.5850	0.6598	0.6202	0.4	0.375	0.3871	0.8852	0.9	0.8926	0.25	0.2812	0.2647	0.56	0.4828	0.5185	0.5366	0.6561
0.005	26.0	3250	0.4058	0.6292	0.6715	0.6496	0.9355	0.5821	0.6433	0.6112	0.3939	0.4062	0.4	0.875	0.8944	0.8846	0.2857	0.25	0.2667	0.5357	0.5172	0.5263	0.5378	0.6494
0.005	27.0	3375	0.3980	0.6262	0.6807	0.6523	0.9369	0.5743	0.6536	0.6114	0.4074	0.3438	0.3729	0.8798	0.8944	0.8871	0.3333	0.375	0.3529	0.5769	0.5172	0.5455	0.5539	0.6533
0.0031	28.0	3500	0.4100	0.6305	0.6755	0.6522	0.9351	0.5762	0.6392	0.6061	0.4074	0.3438	0.3729	0.8962	0.9111	0.9036	0.3421	0.4062	0.3714	0.5385	0.4828	0.5091	0.5526	0.6533
0.0031	29.0	3625	0.4050	0.6383	0.6939	0.6650	0.9388	0.5916	0.6660	0.6266	0.44	0.3438	0.3860	0.8913	0.9111	0.9011	0.3095	0.4062	0.3514	0.5556	0.5172	0.5357	0.5601	0.6665
0.0031	30.0	3750	0.4111	0.6348	0.6741	0.6539	0.9367	0.5819	0.6371	0.6083	0.4138	0.375	0.3934	0.8962	0.9111	0.9036	0.3243	0.375	0.3478	0.56	0.4828	0.5185	0.5543	0.6549
0.0031	31.0	3875	0.4074	0.6349	0.6768	0.6552	0.9381	0.5832	0.6433	0.6118	0.3846	0.3125	0.3448	0.8962	0.9111	0.9036	0.3333	0.4062	0.3662	0.56	0.4828	0.5185	0.5490	0.6559
0.002	32.0	4000	0.4086	0.6421	0.6794	0.6603	0.9379	0.5843	0.6433	0.6124	0.4138	0.375	0.3934	0.9016	0.9167	0.9091	0.375	0.375	0.375	0.5833	0.4828	0.5283	0.5636	0.6603
0.002	33.0	4125	0.4174	0.6378	0.6900	0.6629	0.9369	0.5847	0.6619	0.6209	0.4074	0.3438	0.3729	0.9022	0.9222	0.9121	0.3235	0.3438	0.3333	0.5385	0.4828	0.5091	0.5497	0.6632
0.002	34.0	4250	0.4131	0.6431	0.6847	0.6633	0.9379	0.5881	0.6536	0.6191	0.4138	0.375	0.3934	0.8962	0.9111	0.9036	0.375	0.375	0.375	0.5833	0.4828	0.5283	0.5639	0.6634
0.002	35.0	4375	0.4142	0.6452	0.6860	0.6650	0.9380	0.5911	0.6557	0.6217	0.4138	0.375	0.3934	0.8962	0.9111	0.9036	0.375	0.375	0.375	0.5833	0.4828	0.5283	0.5644	0.6650

Framework versions

Transformers 4.35.2
Pytorch 2.1.0+cu121
Datasets 2.15.0
Tokenizers 0.15.0

csNoHug/roberta-base-finetuned-ner-cadec-no-iob

roberta-base-finetuned-ner-cadec-no-iob

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种