learning_rate: 3e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Use adamw_torch 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: 0.1
num_epochs: 40

Training Loss	Epoch	Step	Validation Loss
3.7299	1.0	124	0.9316
3.6992	2.0	248	0.9297
3.7034	3.0	372	0.9283
3.6984	4.0	496	0.9263
3.6780	5.0	620	0.9270
3.6293	6.0	744	0.9261
3.6348	7.0	868	0.9254
3.6281	8.0	992	0.9259
3.5982	9.0	1116	0.9255
3.6037	10.0	1240	0.9253
3.5761	11.0	1364	0.9244
3.5962	12.0	1488	0.9244
3.6120	13.0	1612	0.9243
3.5741	14.0	1736	0.9240
3.5831	15.0	1860	0.9242
3.5802	16.0	1984	0.9239
3.5810	17.0	2108	0.9241
3.5825	18.0	2232	0.9239
3.6138	19.0	2356	0.9239
3.6049	20.0	2480	0.9241
3.5821	21.0	2604	0.9238
3.6078	22.0	2728	0.9239
3.6065	23.0	2852	0.9239
3.5850	24.0	2976	0.9240
3.5944	25.0	3100	0.9240
3.5877	26.0	3224	0.9240

SociauxLing/modernbert-CGEdit-AAE_final

modernbert-CGEdit-AAE_final