scenario-non-kd-scr-ner-half-mdeberta_data-univner_full66

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3292
• Precision: 0.6067
• Recall: 0.6025
• F1: 0.6046
• Accuracy: 0.9613

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: 32
• eval_batch_size: 32
• seed: 66
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.3569	0.2910	500	0.2788	0.2833	0.1239	0.1724	0.9282
0.2296	0.5821	1000	0.2136	0.3764	0.2307	0.2861	0.9387
0.1733	0.8731	1500	0.1773	0.4309	0.3854	0.4069	0.9475
0.1356	1.1641	2000	0.1639	0.4633	0.4594	0.4613	0.9513
0.1114	1.4552	2500	0.1650	0.5250	0.4467	0.4827	0.9539
0.1001	1.7462	3000	0.1532	0.5105	0.5159	0.5132	0.9559
0.091	2.0373	3500	0.1532	0.5475	0.5320	0.5396	0.9577
0.0659	2.3283	4000	0.1539	0.5492	0.5379	0.5435	0.9580
0.0675	2.6193	4500	0.1489	0.5543	0.5601	0.5572	0.9596
0.0665	2.9104	5000	0.1481	0.5718	0.5755	0.5737	0.9602
0.0469	3.2014	5500	0.1567	0.5695	0.6077	0.5880	0.9600
0.0446	3.4924	6000	0.1607	0.5835	0.5699	0.5766	0.9605
0.0441	3.7835	6500	0.1667	0.5901	0.5588	0.5740	0.9602
0.0414	4.0745	7000	0.1663	0.5764	0.5920	0.5841	0.9606
0.0298	4.3655	7500	0.1695	0.5978	0.5914	0.5946	0.9614
0.0315	4.6566	8000	0.1759	0.5871	0.5910	0.5890	0.9605
0.031	4.9476	8500	0.1725	0.6039	0.5904	0.5971	0.9612
0.0225	5.2386	9000	0.1863	0.5928	0.5937	0.5932	0.9610
0.0218	5.5297	9500	0.1899	0.5953	0.5813	0.5882	0.9607
0.0216	5.8207	10000	0.1881	0.5981	0.5833	0.5906	0.9606
0.0199	6.1118	10500	0.2041	0.5952	0.5819	0.5885	0.9606
0.0155	6.4028	11000	0.2090	0.6155	0.5800	0.5972	0.9613
0.0156	6.6938	11500	0.2053	0.6014	0.5992	0.6003	0.9615
0.0173	6.9849	12000	0.2094	0.5847	0.5869	0.5858	0.9601
0.0112	7.2759	12500	0.2198	0.6023	0.5806	0.5912	0.9606
0.0115	7.5669	13000	0.2210	0.5929	0.5876	0.5902	0.9606
0.0122	7.8580	13500	0.2210	0.5915	0.6133	0.6022	0.9605
0.0104	8.1490	14000	0.2232	0.6016	0.5898	0.5957	0.9605
0.0081	8.4400	14500	0.2360	0.6254	0.5855	0.6048	0.9614
0.0092	8.7311	15000	0.2404	0.6114	0.5776	0.5940	0.9612
0.0093	9.0221	15500	0.2386	0.6090	0.5957	0.6023	0.9613
0.0074	9.3132	16000	0.2434	0.6172	0.5855	0.6009	0.9611
0.0067	9.6042	16500	0.2465	0.6071	0.5889	0.5979	0.9614
0.007	9.8952	17000	0.2531	0.6012	0.5822	0.5915	0.9609
0.0059	10.1863	17500	0.2545	0.6183	0.5683	0.5922	0.9611
0.0057	10.4773	18000	0.2547	0.6134	0.5809	0.5967	0.9609
0.0051	10.7683	18500	0.2634	0.6064	0.5812	0.5935	0.9611
0.005	11.0594	19000	0.2639	0.5916	0.5993	0.5954	0.9607
0.0044	11.3504	19500	0.2598	0.5986	0.5963	0.5975	0.9611
0.0045	11.6414	20000	0.2649	0.5961	0.6159	0.6059	0.9613
0.0046	11.9325	20500	0.2745	0.6138	0.5737	0.5930	0.9612
0.004	12.2235	21000	0.2676	0.5997	0.6087	0.6042	0.9613
0.0036	12.5146	21500	0.2731	0.6208	0.5910	0.6055	0.9620
0.0038	12.8056	22000	0.2794	0.6005	0.5835	0.5919	0.9608
0.0034	13.0966	22500	0.2774	0.6104	0.5891	0.5996	0.9613
0.0025	13.3877	23000	0.2847	0.5956	0.5897	0.5926	0.9608
0.003	13.6787	23500	0.2917	0.6167	0.5690	0.5919	0.9611
0.0031	13.9697	24000	0.2837	0.5991	0.5952	0.5971	0.9608
0.0025	14.2608	24500	0.2858	0.6190	0.5843	0.6012	0.9612
0.0024	14.5518	25000	0.2962	0.6030	0.5807	0.5917	0.9607
0.0029	14.8428	25500	0.2932	0.6159	0.5791	0.5970	0.9611
0.0022	15.1339	26000	0.2960	0.6051	0.6008	0.6030	0.9607
0.0019	15.4249	26500	0.3045	0.6204	0.5770	0.5979	0.9608
0.002	15.7159	27000	0.3014	0.6107	0.5833	0.5967	0.9607
0.0019	16.0070	27500	0.3113	0.6120	0.5882	0.5999	0.9610
0.0014	16.2980	28000	0.3122	0.6070	0.5853	0.5960	0.9607
0.0018	16.5891	28500	0.3091	0.6306	0.5820	0.6053	0.9616
0.0019	16.8801	29000	0.3152	0.6110	0.5812	0.5957	0.9606
0.0017	17.1711	29500	0.3148	0.6291	0.5856	0.6066	0.9616
0.0018	17.4622	30000	0.3091	0.6161	0.5923	0.6039	0.9611
0.0012	17.7532	30500	0.3096	0.6152	0.5855	0.6000	0.9608
0.0013	18.0442	31000	0.3255	0.6172	0.5764	0.5961	0.9610
0.0013	18.3353	31500	0.3189	0.6268	0.5716	0.5979	0.9609
0.0012	18.6263	32000	0.3136	0.6151	0.5918	0.6032	0.9612
0.0015	18.9173	32500	0.3192	0.6093	0.5851	0.5969	0.9609
0.0009	19.2084	33000	0.3219	0.6202	0.5829	0.6010	0.9613
0.0011	19.4994	33500	0.3212	0.6100	0.5969	0.6034	0.9613
0.0009	19.7905	34000	0.3303	0.6381	0.5592	0.5961	0.9611
0.001	20.0815	34500	0.3240	0.6215	0.5829	0.6016	0.9614
0.0009	20.3725	35000	0.3257	0.6243	0.5862	0.6047	0.9616
0.0011	20.6636	35500	0.3319	0.6162	0.5855	0.6004	0.9613
0.0009	20.9546	36000	0.3296	0.6142	0.5835	0.5984	0.9614
0.0006	21.2456	36500	0.3336	0.6133	0.5895	0.6012	0.9613
0.0006	21.5367	37000	0.3373	0.6128	0.5905	0.6015	0.9610
0.0011	21.8277	37500	0.3319	0.6224	0.5888	0.6051	0.9617
0.0007	22.1187	38000	0.3396	0.6140	0.5812	0.5971	0.9611
0.0006	22.4098	38500	0.3337	0.6126	0.5959	0.6041	0.9615
0.0008	22.7008	39000	0.3351	0.6187	0.5928	0.6055	0.9614
0.001	22.9919	39500	0.3292	0.6067	0.6025	0.6046	0.9613

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1