learning_rate: 5e-05
train_batch_size: 2
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
gradient_accumulation_steps: 10
total_train_batch_size: 20
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 10
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Mse	Mae	R2	Spearman Correlation
77.9138	1.0	97	0.3739	0.4525	0.5424	-1.3457	-0.0657
71.0882	2.0	194	0.3739	0.4525	0.5424	-1.3457	-0.0657
53.8100	3.0	291	0.3739	0.4525	0.5424	-1.3457	-0.0657
69.5770	4.0	388	0.3739	0.4525	0.5424	-1.3457	-0.0657
80.0113	5.0	485	0.3739	0.4525	0.5424	-1.3457	-0.0657
69.8547	6.0	582	0.3739	0.4525	0.5424	-1.3457	-0.0657
57.1770	7.0	679	0.3739	0.4525	0.5424	-1.3457	-0.0657
65.3238	8.0	776	0.3739	0.4525	0.5424	-1.3457	-0.0657
73.7931	9.0	873	0.3739	0.4525	0.5424	-1.3457	-0.0657
74.0995	10.0	970	0.3739	0.4525	0.5424	-1.3457	-0.0657

athirorg/USS-reward-model-hybrid-WRS

USS-reward-model-hybrid-WRS