learning_rate: 0.0001
train_batch_size: 16
eval_batch_size: 16
seed: 42
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
lr_scheduler_warmup_steps: 50
num_epochs: 15
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Topology Accuracy	Service Accuracy	Combined Accuracy
0.6178	1.0	96	0.6028	0.9805	0.7012	0.8408
0.5546	2.0	192	0.5229	0.9766	0.7109	0.8438
0.4255	3.0	288	0.4324	0.9805	0.9023	0.9414
0.3934	4.0	384	0.3546	0.9805	0.9668	0.9736
0.3195	5.0	480	0.3477	0.9844	0.9648	0.9746
0.3052	6.0	576	0.3458	0.9863	0.9746	0.9805
0.3139	7.0	672	0.3529	0.9863	0.9688	0.9775
0.3049	8.0	768	0.3533	0.9844	0.9707	0.9775
0.3145	9.0	864	0.3517	0.9863	0.9648	0.9756
0.3025	10.0	960	0.3462	0.9883	0.9668	0.9775
0.3031	11.0	1056	0.3420	0.9902	0.9707	0.9805
0.3008	12.0	1152	0.3417	0.9883	0.9727	0.9805
0.3184	13.0	1248	0.3418	0.9883	0.9746	0.9814
0.2996	14.0	1344	0.3404	0.9883	0.9746	0.9814
0.3023	15.0	1440	0.3406	0.9883	0.9746	0.9814

WaiLwin/jn

jn