learning_rate: 2e-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
19.0462	1.0	97	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0664	2.0	194	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0221	3.0	291	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0183	4.0	388	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0346	5.0	485	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0308	6.0	582	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0287	7.0	679	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0718	8.0	776	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0205	9.0	873	1.8291	0.5939	0.6283	-2.0790	0.0359
19.0156	10.0	970	1.8291	0.5939	0.6283	-2.0790	0.0359

athirorg/USS-reward-model-hybrid-alpha1.0

USS-reward-model-hybrid-alpha1.0