learning_rate: 5e-05
train_batch_size: 4
eval_batch_size: 4
seed: 2025
gradient_accumulation_steps: 4
total_train_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 20
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Subset Accuracy	F1 Macro	F1 Micro	Precision Macro	Recall Macro	Roc Auc
0.4174	0.5737	767	0.2090	0.0	0.0	0.0	0.0	0.0	0.5932
0.1828	1.1473	1534	0.1748	0.1320	0.0803	0.2207	0.1575	0.0620	0.7877
0.1663	1.7210	2301	0.1643	0.2066	0.1822	0.3172	0.4112	0.1480	0.8266
0.1396	2.2947	3068	0.1635	0.2842	0.2466	0.3875	0.4562	0.2071	0.8290
0.1299	2.8684	3835	0.1607	0.2926	0.2628	0.3979	0.4435	0.2153	0.8417
0.0911	3.4420	4602	0.1791	0.3174	0.3135	0.4089	0.4272	0.2707	0.8325
0.0904	4.0157	5369	0.1812	0.3292	0.3212	0.4156	0.3918	0.2816	0.8305
0.0599	4.5894	6136	0.2016	0.3217	0.3415	0.4088	0.4062	0.3074	0.8260
0.0617	5.1631	6903	0.2137	0.3296	0.3433	0.4198	0.4029	0.3095	0.8252
0.041	5.7367	7670	0.2223	0.3171	0.3360	0.4068	0.3899	0.3036	0.8192
0.0454	6.3104	8437	0.2336	0.3213	0.3520	0.4188	0.4061	0.3260	0.8154
0.0308	6.8841	9204	0.2452	0.3134	0.3515	0.4162	0.4116	0.3328	0.8164
0.0297	7.4577	9971	0.2633	0.3069	0.3440	0.4054	0.3786	0.3308	0.8114
0.0235	8.0314	10738	0.2637	0.3115	0.3529	0.4178	0.3922	0.3347	0.8141
0.0169	8.6051	11505	0.2811	0.2937	0.3386	0.4058	0.3737	0.3266	0.8065
0.0175	9.1788	12272	0.2860	0.3065	0.3577	0.4184	0.4005	0.3381	0.8123
0.0119	9.7524	13039	0.2919	0.3083	0.3394	0.4118	0.3805	0.3143	0.8128
0.0118	10.3261	13806	0.3005	0.2954	0.3567	0.4223	0.3666	0.3553	0.8116
0.0089	10.8998	14573	0.3096	0.3034	0.3507	0.4112	0.3758	0.3372	0.8077
0.0079	11.4734	15340	0.3186	0.3031	0.3539	0.4143	0.3900	0.3371	0.8070
0.0064	12.0471	16107	0.3223	0.3042	0.3530	0.4203	0.3824	0.3341	0.8074

DayCardoso/bert-seq-class-values-no-context-cru

bert-seq-class-values-no-context-cru