learning_rate: 5e-05
train_batch_size: 8
eval_batch_size: 8
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
num_epochs: 50
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Bleu
4.7930	1.0	450	2.0807	7.9570
3.0101	2.0	900	1.4390	25.8316
1.4560	3.0	1350	1.2479	27.6850
1.4925	4.0	1800	1.1101	42.0742
0.7328	5.0	2250	1.0813	45.9933
0.5155	6.0	2700	1.1270	43.9782
0.5569	7.0	3150	1.1126	49.1448
0.2561	8.0	3600	1.0845	46.1242
0.2378	9.0	4050	1.1601	52.8584
0.2815	10.0	4500	1.1230	58.7425
0.1808	11.0	4950	1.1898	54.7645
0.1801	12.0	5400	1.1700	61.1318
0.0985	13.0	5850	1.1957	55.3300
0.0944	14.0	6300	1.1710	60.4827

GeethmaYasashwi/Sinhala_Bert_Finetune_BS4_LR5e-5

Sinhala_Bert_Finetune_BS4_LR5e-5