learning_rate: 5e-05
train_batch_size: 4
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
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
num_epochs: 100

Training Loss	Epoch	Step	Validation Loss	Chrf++	Gen Len
25.7837	1.0	2	2.8930	6.6649	188.0
25.7837	2.0	4	2.8969	6.6649	188.0
25.7837	3.0	6	2.8924	7.1427	187.9375
25.7837	4.0	8	2.8937	7.8609	187.9375
25.7837	5.0	10	2.8935	6.4812	188.25
25.7837	6.0	12	2.8967	6.5964	188.0625
25.7837	7.0	14	2.8929	6.8291	188.1875
25.7837	8.0	16	2.8944	6.7433	187.8125
25.7837	9.0	18	2.8912	7.1998	187.8125
25.7837	10.0	20	2.8907	6.9054	188.0625
25.7837	11.0	22	2.8879	7.0101	188.0625
25.7837	12.0	24	2.8870	7.0889	187.9375
25.7837	13.0	26	2.8833	6.8975	188.0625
25.7837	14.0	28	2.8834	5.9333	187.9375
25.7837	15.0	30	2.8808	5.8336	187.9375
25.7837	16.0	32	2.8743	5.8407	187.9375
25.7837	17.0	34	2.8677	6.5105	187.8125
25.7837	18.0	36	2.8610	5.8395	188.0625
25.7837	19.0	38	2.8576	6.8047	188.4375

Tsedeniya/finetuned_byT5base_pan_translit

finetuned_byT5base_pan_translit