learning_rate: 2e-05
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 0.06
num_epochs: 40
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	F1 Micro	F1 Macro
0.3034	1.0	140	0.3127	0.0	0.0
0.2666	2.0	280	0.2452	0.5205	0.2180
0.1898	3.0	420	0.1791	0.6671	0.3777
0.1209	4.0	560	0.1477	0.7124	0.4398
0.1042	5.0	700	0.1304	0.7874	0.5924
0.0823	6.0	840	0.1288	0.7936	0.6420
0.0611	7.0	980	0.1253	0.8005	0.6694
0.0504	8.0	1120	0.1247	0.8037	0.7011
0.0401	9.0	1260	0.1221	0.8105	0.6945
0.0273	10.0	1400	0.1237	0.8145	0.7053
0.0248	11.0	1540	0.1199	0.8171	0.7109
0.0175	12.0	1680	0.1211	0.8159	0.7190
0.0174	13.0	1820	0.1227	0.8233	0.7140
0.0153	14.0	1960	0.1215	0.8272	0.7496
0.0130	15.0	2100	0.1246	0.8259	0.7517
0.0122	16.0	2240	0.1253	0.8220	0.7405
0.0100	17.0	2380	0.1295	0.8235	0.7409
0.0079	18.0	2520	0.1313	0.8213	0.7397
0.0067	19.0	2660	0.1364	0.8173	0.7352
0.0069	20.0	2800	0.1380	0.8221	0.7444

silke-delcart/NLP-final-test

NLP-final-test