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: 10
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 20
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.06
num_epochs: 10
label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Ratio
5.617	0.1626	10	5.2818	0.1471	0.4233	0.0980	0.1518	0.1891
2.9819	0.3252	20	1.8921	0.5462	0.3817	0.3641	0.3655	0.6134
1.4506	0.4878	30	1.3671	0.5378	0.5459	0.5378	0.5165	0.2899
1.112	0.6504	40	0.8974	0.6261	0.6268	0.6261	0.6255	0.4622
0.872	0.8130	50	0.7909	0.7017	0.7320	0.7017	0.6916	0.6807
0.8282	0.9756	60	0.7232	0.7605	0.7614	0.7605	0.7603	0.4706
0.7528	1.1382	70	0.6917	0.7647	0.7654	0.7647	0.7646	0.5252
0.7292	1.3008	80	0.6830	0.7773	0.7789	0.7773	0.7770	0.5378
0.6003	1.4634	90	0.6686	0.7857	0.7968	0.7857	0.7837	0.5966
0.6511	1.6260	100	0.6301	0.8067	0.8071	0.8067	0.8067	0.5168
0.5804	1.7886	110	0.6498	0.7983	0.8004	0.7983	0.7980	0.4580
0.6096	1.9512	120	0.6107	0.8151	0.8152	0.8151	0.8151	0.5084
0.6082	2.1138	130	0.6035	0.8277	0.8283	0.8277	0.8277	0.4790
0.5099	2.2764	140	0.6308	0.8151	0.8155	0.8151	0.8151	0.5168
0.5049	2.4390	150	0.6372	0.8361	0.8381	0.8361	0.8359	0.5378
0.4987	2.6016	160	0.6228	0.8445	0.8446	0.8445	0.8445	0.5042
0.6128	2.7642	170	0.6122	0.8487	0.8488	0.8487	0.8487	0.4916
0.5384	2.9268	180	0.6065	0.8277	0.8346	0.8277	0.8268	0.5714
0.4899	3.0894	190	0.6652	0.8151	0.8195	0.8151	0.8145	0.4412
0.4299	3.2520	200	0.6596	0.8487	0.8512	0.8487	0.8485	0.5420
0.4523	3.4146	210	0.7557	0.8067	0.8110	0.8067	0.8061	0.4412
0.4542	3.5772	220	0.6954	0.8277	0.8283	0.8277	0.8277	0.4790
0.4587	3.7398	230	0.6812	0.8319	0.8323	0.8319	0.8319	0.4832
0.4816	3.9024	240	0.6309	0.8613	0.8634	0.8613	0.8611	0.5378
0.4866	4.0650	250	0.6423	0.8487	0.8503	0.8487	0.8486	0.5336
0.363	4.2276	260	0.6763	0.8445	0.8448	0.8445	0.8445	0.5126
0.399	4.3902	270	0.7227	0.8361	0.8367	0.8361	0.8361	0.4790
0.3862	4.5528	280	0.6777	0.8445	0.8448	0.8445	0.8445	0.5126
0.4815	4.7154	290	0.6559	0.8529	0.8532	0.8529	0.8529	0.5126
0.4548	4.8780	300	0.6757	0.8403	0.8451	0.8403	0.8398	0.4412
0.3675	5.0407	310	0.6526	0.8487	0.8491	0.8487	0.8487	0.5168
0.3626	5.2033	320	0.6815	0.8529	0.8532	0.8529	0.8529	0.5126
0.4256	5.3659	330	0.6904	0.8529	0.8532	0.8529	0.8529	0.4874
0.4515	5.5285	340	0.6561	0.8487	0.8496	0.8487	0.8486	0.5252
0.3661	5.6911	350	0.6681	0.8487	0.8491	0.8487	0.8487	0.5168
0.3792	5.8537	360	0.6740	0.8487	0.8487	0.8487	0.8487	0.5
0.4327	6.0163	370	0.6649	0.8487	0.8487	0.8487	0.8487	0.5
0.3426	6.1789	380	0.6462	0.8487	0.8503	0.8487	0.8486	0.5336
0.3329	6.3415	390	0.6767	0.8529	0.8550	0.8529	0.8527	0.5378
0.415	6.5041	400	0.7001	0.8445	0.8448	0.8445	0.8445	0.4874
0.388	6.6667	410	0.7217	0.8445	0.8457	0.8445	0.8444	0.4706
0.3585	6.8293	420	0.7232	0.8445	0.8457	0.8445	0.8444	0.4706
0.3657	6.9919	430	0.6943	0.8487	0.8496	0.8487	0.8486	0.4748
0.3366	7.1545	440	0.6999	0.8529	0.8536	0.8529	0.8529	0.4790
0.3497	7.3171	450	0.6797	0.8613	0.8614	0.8613	0.8613	0.5042
0.3219	7.4797	460	0.6905	0.8487	0.8496	0.8487	0.8486	0.5252
0.3459	7.6423	470	0.6872	0.8613	0.8614	0.8613	0.8613	0.5042
0.3669	7.8049	480	0.6941	0.8529	0.8536	0.8529	0.8529	0.4790
0.3888	7.9675	490	0.7014	0.8487	0.8496	0.8487	0.8486	0.4748
0.2989	8.1301	500	0.6951	0.8487	0.8488	0.8487	0.8487	0.4916
0.3743	8.2927	510	0.7026	0.8487	0.8488	0.8487	0.8487	0.4916
0.3086	8.4553	520	0.7182	0.8529	0.8532	0.8529	0.8529	0.4874
0.3251	8.6179	530	0.7135	0.8529	0.8532	0.8529	0.8529	0.4874

Framework versions

Transformers 4.40.0
Pytorch 2.2.1+cu121
Datasets 2.19.0
Tokenizers 0.19.1

adriansanz/2504v3

2504v3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

adriansanz/
2504v3