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
lr_scheduler_warmup_steps: 50
num_epochs: 25
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Bleu	Chrf	Meteor	Bertscore F1
2.9664	1.0	189	2.4330	1.1567	17.0736	0.0909	0.7679
1.9794	2.0	378	2.1012	4.5012	23.9044	0.1731	0.795
1.4515	3.0	567	2.0078	6.1795	28.9153	0.2162	0.8106
1.1065	4.0	756	1.9566	10.5487	32.5721	0.2395	0.8193
0.8437	5.0	945	1.9606	9.2165	32.1324	0.2607	0.8245
0.6498	6.0	1134	1.9997	11.145	34.7027	0.2594	0.8293
0.4911	7.0	1323	2.0180	12.3754	36.4547	0.2842	0.8327
0.3768	8.0	1512	2.0067	12.4325	37.4606	0.2817	0.8358
0.2899	9.0	1701	2.0370	15.0857	38.4527	0.3063	0.8408
0.2198	10.0	1890	2.0621	13.5259	37.2397	0.2933	0.8366
0.1704	11.0	2079	2.1010	14.3345	37.9934	0.2929	0.8391
0.1355	12.0	2268	2.1056	14.9335	38.6265	0.3077	0.8418
0.1118	13.0	2457	2.1115	13.5777	38.8908	0.3181	0.8425
0.0906	14.0	2646	2.1289	14.5873	38.5979	0.3099	0.8431
0.0748	15.0	2835	2.1332	14.1024	39.7978	0.3117	0.8422
0.0638	16.0	3024	2.1610	15.3784	39.4667	0.3131	0.8422
0.0571	17.0	3213	2.1501	15.7275	40.414	0.3245	0.8476
0.0519	18.0	3402	2.1482	16.9568	40.3169	0.3142	0.848
0.0468	19.0	3591	2.1651	15.4548	40.654	0.323	0.847
0.0399	20.0	3780	2.1768	15.4357	40.1039	0.3233	0.8499
0.0378	21.0	3969	2.1720	14.9499	40.049	0.3148	0.8481
0.0356	22.0	4158	2.1779	15.3876	40.6863	0.3256	0.8499
0.0329	23.0	4347	2.1900	15.6909	41.2235	0.3275	0.8497
0.0309	24.0	4536	2.1893	14.8779	40.3529	0.319	0.847
0.0300	25.0	4725	2.1860	15.0397	40.4001	0.3216	0.8476

thunderboltc/nllb_ipa_to_bangla_epoch25

nllb_ipa_to_bangla_epoch25