clinico-bsc-bio-ehr-es-finetuned-v2

This model is a fine-tuned version of joheras/bsc-bio-ehr-es-finetuned-clinais-v2 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.9998
Precision: 0.4800
Recall: 0.6452
F1: 0.5505
Accuracy: 0.8566

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: 32
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	25	1.2290	0.0070	0.0137	0.0093	0.5710
No log	2.0	50	0.7436	0.1540	0.1943	0.1718	0.7956
No log	3.0	75	0.6477	0.2407	0.3126	0.2719	0.8209
No log	4.0	100	0.6309	0.3249	0.4467	0.3762	0.8267
No log	5.0	125	0.6076	0.3437	0.4572	0.3924	0.8286
No log	6.0	150	0.6041	0.3445	0.4551	0.3922	0.8328
No log	7.0	175	0.5999	0.3242	0.4382	0.3727	0.8376
No log	8.0	200	0.6016	0.3551	0.5005	0.4154	0.8355
No log	9.0	225	0.6019	0.3628	0.5375	0.4332	0.8523
No log	10.0	250	0.6052	0.3340	0.5259	0.4085	0.8463
No log	11.0	275	0.6518	0.3405	0.5354	0.4163	0.8449
No log	12.0	300	0.6613	0.3851	0.5660	0.4583	0.8500
No log	13.0	325	0.6725	0.3737	0.5702	0.4515	0.8488
No log	14.0	350	0.6770	0.3889	0.5892	0.4685	0.8528
No log	15.0	375	0.6816	0.3742	0.5702	0.4519	0.8471
No log	16.0	400	0.6998	0.3830	0.5808	0.4616	0.8498
No log	17.0	425	0.7434	0.3789	0.5882	0.4609	0.8474
No log	18.0	450	0.7543	0.4176	0.6051	0.4942	0.8504
No log	19.0	475	0.7537	0.4196	0.6114	0.4976	0.8521
0.3481	20.0	500	0.7482	0.4102	0.6199	0.4937	0.8489
0.3481	21.0	525	0.7488	0.4203	0.6241	0.5023	0.8545
0.3481	22.0	550	0.7603	0.4097	0.6325	0.4973	0.8548
0.3481	23.0	575	0.7809	0.4333	0.6177	0.5094	0.8522
0.3481	24.0	600	0.7873	0.4317	0.6272	0.5114	0.8545
0.3481	25.0	625	0.7914	0.4368	0.6241	0.5139	0.8581
0.3481	26.0	650	0.8077	0.4508	0.6241	0.5235	0.8526
0.3481	27.0	675	0.8091	0.4639	0.6304	0.5345	0.8589
0.3481	28.0	700	0.8218	0.4543	0.6346	0.5295	0.8558
0.3481	29.0	725	0.8316	0.4522	0.6241	0.5244	0.8545
0.3481	30.0	750	0.8444	0.4434	0.6325	0.5213	0.8535
0.3481	31.0	775	0.8208	0.4506	0.6209	0.5222	0.8559
0.3481	32.0	800	0.8425	0.4572	0.6262	0.5285	0.8554
0.3481	33.0	825	0.8872	0.4453	0.6272	0.5208	0.8494
0.3481	34.0	850	0.8533	0.4635	0.6167	0.5292	0.8579
0.3481	35.0	875	0.8927	0.4529	0.6188	0.5230	0.8532
0.3481	36.0	900	0.8919	0.4659	0.6199	0.5319	0.8517
0.3481	37.0	925	0.8971	0.4480	0.6272	0.5227	0.8544
0.3481	38.0	950	0.9032	0.4708	0.6378	0.5417	0.8541
0.3481	39.0	975	0.8805	0.4449	0.6315	0.5220	0.8550
0.0239	40.0	1000	0.8927	0.4580	0.6283	0.5298	0.8567
0.0239	41.0	1025	0.9016	0.4906	0.6315	0.5522	0.8587
0.0239	42.0	1050	0.9078	0.4691	0.6336	0.5391	0.8572
0.0239	43.0	1075	0.9275	0.4748	0.6272	0.5405	0.8534
0.0239	44.0	1100	0.9360	0.4559	0.6272	0.528	0.8508
0.0239	45.0	1125	0.9357	0.4645	0.6357	0.5368	0.8568
0.0239	46.0	1150	0.9359	0.4730	0.6294	0.5401	0.8541
0.0239	47.0	1175	0.9388	0.4730	0.6294	0.5401	0.8550
0.0239	48.0	1200	0.9322	0.4832	0.6389	0.5503	0.8593
0.0239	49.0	1225	0.9519	0.4755	0.6262	0.5406	0.8543
0.0239	50.0	1250	0.9503	0.4674	0.6283	0.5360	0.8515
0.0239	51.0	1275	0.9547	0.4773	0.6336	0.5445	0.8520
0.0239	52.0	1300	0.9460	0.4767	0.6367	0.5452	0.8550
0.0239	53.0	1325	0.9537	0.4704	0.6283	0.5380	0.8552
0.0239	54.0	1350	0.9485	0.4759	0.6367	0.5447	0.8553
0.0239	55.0	1375	0.9503	0.4737	0.6367	0.5432	0.8570
0.0239	56.0	1400	0.9585	0.4890	0.6346	0.5524	0.8542
0.0239	57.0	1425	0.9532	0.4864	0.6420	0.5535	0.8572
0.0239	58.0	1450	0.9555	0.4734	0.6399	0.5442	0.8578
0.0239	59.0	1475	0.9683	0.4704	0.6304	0.5388	0.8517
0.009	60.0	1500	0.9736	0.4755	0.6357	0.5441	0.8547
0.009	61.0	1525	0.9656	0.4871	0.6399	0.5532	0.8553
0.009	62.0	1550	0.9735	0.4705	0.6315	0.5392	0.8531
0.009	63.0	1575	0.9700	0.4757	0.6420	0.5465	0.8568
0.009	64.0	1600	0.9749	0.4908	0.6452	0.5575	0.8539
0.009	65.0	1625	0.9815	0.5021	0.6389	0.5623	0.8536
0.009	66.0	1650	0.9829	0.4895	0.6399	0.5547	0.8553
0.009	67.0	1675	0.9950	0.4845	0.6420	0.5522	0.8537
0.009	68.0	1700	0.9822	0.5033	0.6452	0.5655	0.8566
0.009	69.0	1725	0.9747	0.5037	0.6526	0.5685	0.8594
0.009	70.0	1750	0.9825	0.4841	0.6410	0.5516	0.8560
0.009	71.0	1775	0.9839	0.4936	0.6484	0.5605	0.8561
0.009	72.0	1800	0.9855	0.4873	0.6494	0.5568	0.8570
0.009	73.0	1825	0.9845	0.4924	0.6494	0.5601	0.8567
0.009	74.0	1850	0.9842	0.4880	0.6463	0.5561	0.8561
0.009	75.0	1875	0.9892	0.4853	0.6441	0.5535	0.8559
0.009	76.0	1900	0.9899	0.4865	0.6463	0.5551	0.8568
0.009	77.0	1925	0.9861	0.4947	0.6420	0.5588	0.8582
0.009	78.0	1950	0.9875	0.4774	0.6473	0.5495	0.8596
0.009	79.0	1975	0.9827	0.5033	0.6473	0.5663	0.8600
0.0056	80.0	2000	0.9876	0.4734	0.6473	0.5468	0.8602
0.0056	81.0	2025	0.9957	0.4834	0.6463	0.5531	0.8574
0.0056	82.0	2050	0.9883	0.4766	0.6463	0.5486	0.8569
0.0056	83.0	2075	0.9904	0.4903	0.6420	0.5560	0.8567
0.0056	84.0	2100	0.9920	0.4829	0.6410	0.5508	0.8572
0.0056	85.0	2125	0.9817	0.4947	0.6463	0.5604	0.8600
0.0056	86.0	2150	0.9837	0.4900	0.6484	0.5582	0.8601
0.0056	87.0	2175	0.9908	0.4852	0.6420	0.5527	0.8572
0.0056	88.0	2200	0.9873	0.5049	0.6484	0.5677	0.8590
0.0056	89.0	2225	0.9924	0.4843	0.6367	0.5502	0.8560
0.0056	90.0	2250	0.9897	0.4794	0.6378	0.5473	0.8574
0.0056	91.0	2275	0.9956	0.4779	0.6389	0.5468	0.8569
0.0056	92.0	2300	1.0018	0.4829	0.6410	0.5508	0.8567
0.0056	93.0	2325	1.0021	0.4753	0.6410	0.5459	0.8562
0.0056	94.0	2350	0.9968	0.4817	0.6410	0.5501	0.8575
0.0056	95.0	2375	0.9940	0.4799	0.6441	0.5500	0.8584
0.0056	96.0	2400	0.9964	0.4780	0.6420	0.5480	0.8577
0.0056	97.0	2425	1.0001	0.4785	0.6463	0.5499	0.8578
0.0056	98.0	2450	0.9989	0.4797	0.6473	0.5510	0.8582
0.0056	99.0	2475	0.9995	0.4762	0.6452	0.5480	0.8572
0.0043	100.0	2500	0.9998	0.4800	0.6452	0.5505	0.8566

Framework versions

Transformers 4.25.1
Pytorch 1.13.0
Datasets 2.8.0
Tokenizers 0.12.1

joheras/clinico-bsc-bio-ehr-es-finetuned-v2

clinico-bsc-bio-ehr-es-finetuned-v2

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种