scenario-kd-scr-ner-half_data-univner_full55

This model is a fine-tuned version of haryoaw/scenario-TCR-NER_data-univner_full on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.1043
• Precision: 0.5956
• Recall: 0.5386
• F1: 0.5656
• Accuracy: 0.9591

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: 55
• 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
2.8001	0.2910	500	2.4909	0.4352	0.0136	0.0263	0.9245
2.1119	0.5821	1000	2.1519	0.2415	0.1440	0.1804	0.9270
1.9078	0.8731	1500	2.0224	0.2495	0.1339	0.1743	0.9298
1.7762	1.1641	2000	1.9231	0.2552	0.1578	0.1950	0.9318
1.6349	1.4552	2500	1.8712	0.3116	0.2858	0.2982	0.9346
1.5939	1.7462	3000	1.7761	0.3273	0.3079	0.3173	0.9377
1.4775	2.0373	3500	1.7624	0.3492	0.2557	0.2952	0.9385
1.388	2.3283	4000	1.6539	0.3832	0.3331	0.3564	0.9407
1.3144	2.6193	4500	1.6543	0.3350	0.3226	0.3287	0.9379
1.2781	2.9104	5000	1.5514	0.4049	0.3541	0.3778	0.9436
1.1939	3.2014	5500	1.5313	0.4128	0.3630	0.3863	0.9438
1.137	3.4924	6000	1.5061	0.4120	0.3988	0.4053	0.9447
1.1309	3.7835	6500	1.4980	0.4210	0.3520	0.3834	0.9453
1.0593	4.0745	7000	1.4894	0.4453	0.3795	0.4098	0.9470
0.993	4.3655	7500	1.4434	0.4509	0.3988	0.4232	0.9481
0.9673	4.6566	8000	1.4375	0.4632	0.4165	0.4386	0.9489
0.9454	4.9476	8500	1.3964	0.4640	0.4240	0.4431	0.9489
0.8715	5.2386	9000	1.4012	0.4630	0.4288	0.4452	0.9506
0.8343	5.5297	9500	1.3712	0.4531	0.4497	0.4514	0.9498
0.8396	5.8207	10000	1.3719	0.5407	0.4007	0.4603	0.9504
0.8217	6.1118	10500	1.3344	0.5002	0.4810	0.4904	0.9521
0.7583	6.4028	11000	1.3398	0.4861	0.4768	0.4814	0.9513
0.7399	6.6938	11500	1.3444	0.4971	0.4903	0.4937	0.9524
0.7296	6.9849	12000	1.3103	0.4902	0.5073	0.4986	0.9526
0.6707	7.2759	12500	1.3216	0.5082	0.4872	0.4975	0.9536
0.6701	7.5669	13000	1.2793	0.5306	0.4960	0.5127	0.9547
0.6496	7.8580	13500	1.2766	0.5139	0.5031	0.5084	0.9541
0.6243	8.1490	14000	1.2645	0.5363	0.5045	0.5200	0.9555
0.6003	8.4400	14500	1.2612	0.5457	0.4882	0.5154	0.9552
0.5937	8.7311	15000	1.2817	0.5302	0.4822	0.5051	0.9546
0.5766	9.0221	15500	1.2389	0.5441	0.4976	0.5198	0.9556
0.5478	9.3132	16000	1.2498	0.5628	0.4926	0.5254	0.9564
0.5367	9.6042	16500	1.2404	0.5506	0.5233	0.5366	0.9564
0.5329	9.8952	17000	1.2316	0.5385	0.5259	0.5321	0.9556
0.5139	10.1863	17500	1.2420	0.5340	0.5258	0.5298	0.9560
0.4943	10.4773	18000	1.2187	0.5565	0.5259	0.5408	0.9570
0.4918	10.7683	18500	1.2259	0.5625	0.5006	0.5298	0.9564
0.4971	11.0594	19000	1.2147	0.5591	0.5273	0.5428	0.9573
0.4611	11.3504	19500	1.2211	0.5714	0.4996	0.5331	0.9568
0.4613	11.6414	20000	1.2157	0.5513	0.5243	0.5375	0.9565
0.4601	11.9325	20500	1.2139	0.5717	0.5120	0.5402	0.9569
0.4432	12.2235	21000	1.1935	0.5585	0.5445	0.5514	0.9575
0.4319	12.5146	21500	1.1940	0.5683	0.5483	0.5581	0.9580
0.4339	12.8056	22000	1.1763	0.5703	0.5439	0.5568	0.9575
0.4203	13.0966	22500	1.1844	0.5693	0.5334	0.5508	0.9577
0.4071	13.3877	23000	1.2131	0.5818	0.5146	0.5462	0.9573
0.4055	13.6787	23500	1.1782	0.5796	0.5263	0.5517	0.9575
0.4062	13.9697	24000	1.1744	0.5781	0.5312	0.5537	0.9579
0.3878	14.2608	24500	1.1585	0.5748	0.5532	0.5638	0.9579
0.3894	14.5518	25000	1.1790	0.5823	0.5282	0.5539	0.9581
0.3791	14.8428	25500	1.1633	0.5806	0.5304	0.5544	0.9579
0.3776	15.1339	26000	1.1769	0.5894	0.5259	0.5559	0.9584
0.3633	15.4249	26500	1.1719	0.5721	0.5507	0.5612	0.9583
0.364	15.7159	27000	1.1530	0.5744	0.5488	0.5613	0.9583
0.3627	16.0070	27500	1.1526	0.5711	0.5532	0.5620	0.9581
0.3475	16.2980	28000	1.1648	0.5943	0.5230	0.5564	0.9583
0.3464	16.5891	28500	1.1688	0.5796	0.5132	0.5444	0.9580
0.3466	16.8801	29000	1.1459	0.5936	0.5268	0.5582	0.9583
0.3405	17.1711	29500	1.1623	0.5956	0.5299	0.5608	0.9578
0.3305	17.4622	30000	1.1574	0.5856	0.5282	0.5554	0.9577
0.3359	17.7532	30500	1.1441	0.5861	0.5468	0.5658	0.9588
0.3266	18.0442	31000	1.1436	0.5934	0.5410	0.5660	0.9586
0.3211	18.3353	31500	1.1442	0.5754	0.5438	0.5592	0.9583
0.3178	18.6263	32000	1.1365	0.5907	0.5468	0.5679	0.9588
0.3218	18.9173	32500	1.1396	0.5866	0.5448	0.5649	0.9584
0.3147	19.2084	33000	1.1456	0.6075	0.5175	0.5589	0.9579
0.3088	19.4994	33500	1.1359	0.5733	0.5301	0.5509	0.9581
0.3096	19.7905	34000	1.1376	0.6005	0.5317	0.5640	0.9587
0.3071	20.0815	34500	1.1227	0.5936	0.5618	0.5773	0.9591
0.3046	20.3725	35000	1.1187	0.5972	0.5428	0.5687	0.9586
0.2978	20.6636	35500	1.1395	0.5925	0.5442	0.5673	0.9587
0.2999	20.9546	36000	1.1410	0.5992	0.5468	0.5718	0.9591
0.2916	21.2456	36500	1.1224	0.5909	0.5464	0.5678	0.9587
0.2917	21.5367	37000	1.1380	0.6075	0.5305	0.5664	0.9585
0.2923	21.8277	37500	1.1057	0.5955	0.5549	0.5745	0.9593
0.2898	22.1187	38000	1.1172	0.6055	0.5432	0.5727	0.9587
0.2833	22.4098	38500	1.1302	0.6125	0.5249	0.5653	0.9584
0.2827	22.7008	39000	1.1194	0.6053	0.5462	0.5742	0.9590
0.2873	22.9919	39500	1.1203	0.5920	0.5418	0.5658	0.9586
0.2792	23.2829	40000	1.1254	0.6032	0.5389	0.5692	0.9586
0.2771	23.5739	40500	1.1243	0.5987	0.5301	0.5623	0.9586
0.2788	23.8650	41000	1.1294	0.6103	0.5083	0.5546	0.9582
0.2754	24.1560	41500	1.1016	0.5991	0.5498	0.5734	0.9596
0.2742	24.4470	42000	1.1133	0.5986	0.5496	0.5730	0.9591
0.2716	24.7381	42500	1.1239	0.6167	0.5198	0.5642	0.9585
0.2716	25.0291	43000	1.1182	0.5941	0.5416	0.5666	0.9588
0.268	25.3201	43500	1.1049	0.6014	0.5461	0.5724	0.9589
0.2673	25.6112	44000	1.1126	0.6006	0.5298	0.5630	0.9586
0.2667	25.9022	44500	1.1043	0.5956	0.5386	0.5656	0.9591

Framework versions

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