learning_rate: 1e-05
train_batch_size: 2
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 2
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
1.4159	0.0804	500	1.2113	26.8618	13.7400	22.6242	22.5895
1.3831	0.1608	1000	1.1629	27.1682	14.2922	22.9340	22.9407
1.4429	0.2412	1500	1.1740	26.3174	13.3691	22.1695	22.1162
1.449	0.3216	2000	1.1724	26.2449	13.2833	22.0983	22.0470
1.4263	0.4020	2500	1.1725	26.2259	13.2458	22.0971	22.0603
1.4482	0.4824	3000	1.1725	26.2291	13.2776	22.1100	22.0645
1.4241	0.5628	3500	1.1725	26.2545	13.2396	22.1039	22.0575
1.4162	0.6432	4000	1.1725	26.2714	13.2853	22.1341	22.1005
1.4222	0.7236	4500	1.1725	26.2655	13.2861	22.1245	22.0911
1.3981	0.8040	5000	1.1725	26.2334	13.2644	22.1037	22.0614
1.4436	0.8844	5500	1.1726	26.2563	13.2686	22.1135	22.0862
1.4375	0.9648	6000	1.1725	26.2445	13.2861	22.1145	22.0876
1.4218	1.0452	6500	1.1726	26.2385	13.2984	22.1388	22.1080
1.4205	1.1256	7000	1.1725	26.2472	13.2703	22.1307	22.0987
1.4015	1.2060	7500	1.1725	26.1918	13.2530	22.0918	22.0415
1.4307	1.2864	8000	1.1725	26.2286	13.2684	22.1133	22.0767
1.4436	1.3668	8500	1.1725	26.1782	13.2320	22.0806	22.0392
1.4147	1.4472	9000	1.1725	26.2300	13.2703	22.1014	22.0480
1.4059	1.5276	9500	1.1725	26.2352	13.2800	22.1248	22.0848
1.3891	1.608	10000	1.1725	26.2352	13.2800	22.1248	22.0848
0.0	1.6884	10500	nan	0.0	0.0	0.0	0.0
0.0	1.7688	11000	nan	0.0	0.0	0.0	0.0
0.0	1.8492	11500	nan	0.0	0.0	0.0	0.0
0.0	1.9296	12000	nan	0.0	0.0	0.0	0.0

KalvinPhan/t5-base-cnn

t5-base-cnn