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
14.2826	1.0	211	1.3248	8.4925	134.2339
7.6020	2.0	422	0.9370	16.4972	61.7202
5.1466	3.0	633	0.5313	38.2669	36.344
2.8768	4.0	844	0.3373	55.6968	48.4174
1.8427	5.0	1055	0.2494	69.245	39.2936
1.3324	6.0	1266	0.2731	69.9864	37.3486
0.9895	7.0	1477	0.2184	75.2135	37.5505
0.7888	8.0	1688	0.2120	74.4282	36.844
0.6257	9.0	1899	0.2389	71.5283	46.5046
0.5513	10.0	2110	0.2137	75.6952	39.8028
0.4829	11.0	2321	0.2185	73.6075	42.5642
0.4147	12.0	2532	0.2494	75.8968	37.3716
0.3852	13.0	2743	0.2306	76.2278	38.367
0.3486	14.0	2954	0.2250	76.6194	36.922
0.3119	15.0	3165	0.2248	78.2788	37.4679
0.3026	16.0	3376	0.3328	75.0652	37.1927
0.2778	17.0	3587	0.2351	75.3358	36.5963
0.2680	18.0	3798	0.2864	74.4924	36.6193

Tsedeniya/finetuned_byT5base_ain

finetuned_byT5base_ain