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

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3518
• Precision: 0.5316
• Recall: 0.5425
• F1: 0.5370
• Accuracy: 0.9565

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
0.3589	0.2910	500	0.2912	0.4556	0.0998	0.1638	0.9285
0.2772	0.5821	1000	0.2813	0.4297	0.1296	0.1991	0.9314
0.2452	0.8731	1500	0.2362	0.3233	0.2021	0.2487	0.9349
0.2225	1.1641	2000	0.2304	0.3518	0.2447	0.2886	0.9377
0.2028	1.4552	2500	0.2316	0.3541	0.2378	0.2845	0.9385
0.1966	1.7462	3000	0.2155	0.3847	0.2730	0.3194	0.9401
0.1841	2.0373	3500	0.2109	0.3783	0.3202	0.3468	0.9408
0.1651	2.3283	4000	0.2105	0.3801	0.3388	0.3582	0.9404
0.1644	2.6193	4500	0.2056	0.3732	0.3506	0.3616	0.9417
0.1635	2.9104	5000	0.1966	0.3919	0.3705	0.3809	0.9429
0.1441	3.2014	5500	0.2022	0.4122	0.3985	0.4052	0.9432
0.1418	3.4924	6000	0.2040	0.4160	0.3760	0.3950	0.9443
0.1363	3.7835	6500	0.2053	0.4157	0.3682	0.3905	0.9446
0.1367	4.0745	7000	0.2032	0.4182	0.3929	0.4051	0.9448
0.1203	4.3655	7500	0.2053	0.4173	0.4007	0.4088	0.9448
0.1222	4.6566	8000	0.2040	0.4337	0.3936	0.4127	0.9459
0.1215	4.9476	8500	0.1982	0.4271	0.4227	0.4249	0.9455
0.1072	5.2386	9000	0.2090	0.4375	0.4162	0.4266	0.9470
0.1035	5.5297	9500	0.2049	0.4449	0.4233	0.4338	0.9477
0.0988	5.8207	10000	0.1970	0.4446	0.4558	0.4501	0.9469
0.0935	6.1118	10500	0.2087	0.4576	0.4489	0.4532	0.9485
0.0847	6.4028	11000	0.2098	0.4410	0.4639	0.4521	0.9487
0.0811	6.6938	11500	0.2072	0.4662	0.4699	0.4680	0.9506
0.0828	6.9849	12000	0.1986	0.4947	0.4657	0.4798	0.9510
0.0681	7.2759	12500	0.2098	0.4742	0.4797	0.4769	0.9515
0.0664	7.5669	13000	0.2018	0.4830	0.4887	0.4858	0.9511
0.0674	7.8580	13500	0.2066	0.4954	0.5084	0.5018	0.9532
0.0621	8.1490	14000	0.2088	0.4737	0.5086	0.4905	0.9513
0.0532	8.4400	14500	0.2197	0.4995	0.4777	0.4883	0.9528
0.0544	8.7311	15000	0.2195	0.5120	0.4793	0.4951	0.9528
0.0558	9.0221	15500	0.2174	0.4953	0.5044	0.4998	0.9533
0.0454	9.3132	16000	0.2241	0.5061	0.5095	0.5078	0.9536
0.0458	9.6042	16500	0.2215	0.5058	0.5227	0.5141	0.9540
0.0451	9.8952	17000	0.2181	0.4940	0.5200	0.5066	0.9525
0.0399	10.1863	17500	0.2318	0.5085	0.5194	0.5139	0.9538
0.0375	10.4773	18000	0.2378	0.5108	0.5240	0.5173	0.9541
0.0378	10.7683	18500	0.2312	0.5118	0.5255	0.5185	0.9543
0.0376	11.0594	19000	0.2445	0.5006	0.5074	0.5040	0.9540
0.0338	11.3504	19500	0.2455	0.5081	0.5120	0.5101	0.9543
0.0326	11.6414	20000	0.2442	0.5108	0.5321	0.5212	0.9546
0.0318	11.9325	20500	0.2495	0.5168	0.5171	0.5169	0.9550
0.0289	12.2235	21000	0.2487	0.5113	0.5350	0.5229	0.9550
0.0278	12.5146	21500	0.2522	0.5050	0.5263	0.5154	0.9543
0.0277	12.8056	22000	0.2608	0.5221	0.5138	0.5179	0.9548
0.0263	13.0966	22500	0.2561	0.5133	0.5269	0.5200	0.9551
0.024	13.3877	23000	0.2631	0.5196	0.5258	0.5227	0.9547
0.0246	13.6787	23500	0.2628	0.5110	0.5527	0.5311	0.9551
0.0241	13.9697	24000	0.2735	0.5161	0.5260	0.5210	0.9552
0.021	14.2608	24500	0.2737	0.5224	0.5256	0.5240	0.9551
0.0201	14.5518	25000	0.2743	0.5246	0.5360	0.5302	0.9554
0.0208	14.8428	25500	0.2776	0.5180	0.5266	0.5222	0.9552
0.0201	15.1339	26000	0.2801	0.5065	0.5370	0.5213	0.9549
0.018	15.4249	26500	0.2770	0.5168	0.5335	0.5250	0.9550
0.0176	15.7159	27000	0.2875	0.5185	0.5324	0.5253	0.9551
0.0177	16.0070	27500	0.2861	0.5267	0.5321	0.5294	0.9556
0.0148	16.2980	28000	0.2860	0.5079	0.5442	0.5254	0.9549
0.0156	16.5891	28500	0.2953	0.5188	0.5380	0.5282	0.9552
0.0165	16.8801	29000	0.2928	0.5261	0.5333	0.5297	0.9557
0.0135	17.1711	29500	0.2981	0.5171	0.5396	0.5281	0.9554
0.0142	17.4622	30000	0.3062	0.5269	0.5164	0.5216	0.9554
0.0134	17.7532	30500	0.2947	0.5211	0.5418	0.5312	0.9555
0.0134	18.0442	31000	0.3045	0.5188	0.5426	0.5305	0.9559
0.012	18.3353	31500	0.3070	0.5236	0.5380	0.5307	0.9558
0.0123	18.6263	32000	0.3071	0.5409	0.5328	0.5368	0.9567
0.0117	18.9173	32500	0.3094	0.5265	0.5357	0.5311	0.9560
0.0108	19.2084	33000	0.3167	0.5344	0.5305	0.5325	0.9565
0.0111	19.4994	33500	0.3162	0.5182	0.5302	0.5241	0.9556
0.011	19.7905	34000	0.3152	0.5243	0.5377	0.5309	0.9557
0.0106	20.0815	34500	0.3241	0.5354	0.5200	0.5276	0.9562
0.0094	20.3725	35000	0.3240	0.5223	0.5288	0.5255	0.9560
0.0094	20.6636	35500	0.3271	0.5293	0.5322	0.5308	0.9563
0.0099	20.9546	36000	0.3219	0.5256	0.5334	0.5295	0.9559
0.0085	21.2456	36500	0.3223	0.5245	0.5429	0.5335	0.9560
0.0081	21.5367	37000	0.3308	0.5170	0.5340	0.5254	0.9558
0.0095	21.8277	37500	0.3292	0.5333	0.5294	0.5313	0.9564
0.008	22.1187	38000	0.3326	0.5270	0.5416	0.5342	0.9563
0.007	22.4098	38500	0.3306	0.5252	0.5473	0.5360	0.9563
0.0083	22.7008	39000	0.3301	0.5354	0.5396	0.5375	0.9565
0.0079	22.9919	39500	0.3268	0.5357	0.5421	0.5389	0.9562
0.0072	23.2829	40000	0.3383	0.5367	0.5311	0.5339	0.9563
0.0068	23.5739	40500	0.3349	0.5281	0.5392	0.5336	0.9562
0.0069	23.8650	41000	0.3383	0.5280	0.5408	0.5343	0.9563
0.0073	24.1560	41500	0.3390	0.5217	0.5436	0.5324	0.9563
0.0057	24.4470	42000	0.3395	0.5279	0.5311	0.5295	0.9560
0.0064	24.7381	42500	0.3420	0.5403	0.5295	0.5349	0.9563
0.0065	25.0291	43000	0.3436	0.5372	0.5348	0.5360	0.9565
0.0053	25.3201	43500	0.3444	0.5259	0.5399	0.5328	0.9562
0.0058	25.6112	44000	0.3475	0.5160	0.5367	0.5261	0.9556
0.0061	25.9022	44500	0.3479	0.5393	0.5344	0.5369	0.9566
0.0051	26.1932	45000	0.3435	0.5266	0.5418	0.5341	0.9559
0.0055	26.4843	45500	0.3440	0.5282	0.5419	0.5350	0.9562
0.005	26.7753	46000	0.3466	0.5287	0.5423	0.5354	0.9564
0.0058	27.0664	46500	0.3470	0.5308	0.5490	0.5398	0.9565
0.0052	27.3574	47000	0.3506	0.5343	0.5379	0.5361	0.9564
0.0049	27.6484	47500	0.3475	0.5276	0.5473	0.5373	0.9563
0.0052	27.9395	48000	0.3496	0.5276	0.5483	0.5377	0.9565
0.0049	28.2305	48500	0.3507	0.5327	0.5422	0.5374	0.9564
0.0049	28.5215	49000	0.3528	0.5363	0.5399	0.5381	0.9565
0.0052	28.8126	49500	0.3516	0.5382	0.5385	0.5383	0.9565
0.0042	29.1036	50000	0.3499	0.5330	0.5454	0.5391	0.9565
0.0045	29.3946	50500	0.3514	0.5343	0.5389	0.5366	0.9565
0.0048	29.6857	51000	0.3517	0.5316	0.5418	0.5367	0.9564
0.0043	29.9767	51500	0.3518	0.5316	0.5425	0.5370	0.9565

Framework versions

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