learning_rate: 1e-05
train_batch_size: 4
eval_batch_size: 4
seed: 42
distributed_type: multi-GPU
num_devices: 2
total_train_batch_size: 8
total_eval_batch_size: 8
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 10
mixed_precision_training: Native AMP
label_smoothing_factor: 0.1

Training Loss	Epoch	Step	Validation Loss	Micro F1
1.3184	1.0	1083	0.9248	0.7986
0.858	2.0	2166	0.8338	0.8424
0.6955	3.0	3249	0.8040	0.8561
0.5752	4.0	4332	0.8284	0.8463
0.5152	5.0	5415	0.8289	0.8502
0.4921	6.0	6498	0.8361	0.8424
0.4807	7.0	7581	0.8297	0.8502
0.4752	8.0	8664	0.8271	0.8561
0.4717	9.0	9747	0.8349	0.8476
0.4696	10.0	10830	0.8325	0.8502

AllanK24/modernBERT-Math-Classifier

modernBERT-Math-Classifier