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
7.5124	1.9941	420	0.8551	19.7976	44.1697
2.7690	3.9845	840	0.3240	63.3525	38.6606
1.2601	5.9750	1260	0.2289	72.8755	37.2156
0.7756	7.9655	1680	0.2633	73.7389	37.6376
0.5451	9.9560	2100	0.2038	76.8909	37.9312
0.4182	11.9465	2520	0.2002	80.2369	37.3853
0.3393	13.9370	2940	0.2037	79.5471	39.5917
0.2962	15.9275	3360	0.1993	80.5978	38.2156
0.2705	17.9180	3780	0.2330	79.3043	37.5872
0.3150	19.9084	4200	0.2167	78.4446	38.1193
0.2663	21.8989	4620	0.2331	79.1888	40.1284

Tsedeniya/finetuned_byT5base_ain_joint

finetuned_byT5base_ain_joint