learning_rate: 5e-05
train_batch_size: 16
eval_batch_size: 16
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
num_epochs: 20

Training Loss	Epoch	Step	Validation Loss	Bleu	Ter	Meteor	Chrf
3.4807	1.0	81	1.9528	10.5771	75.3676	0.3663	37.5181
1.3969	2.0	162	1.5794	17.0487	62.8676	0.4594	45.3205
0.9809	3.0	243	1.2933	21.3958	59.4363	0.4789	48.8722
0.5687	4.0	324	1.2516	23.5237	59.5588	0.5025	51.5069
0.3665	5.0	405	1.2780	23.8626	68.0147	0.5298	52.4311
0.2579	6.0	486	1.2683	29.1979	56.9853	0.5391	54.3828
0.1656	7.0	567	1.2798	29.6347	54.2892	0.5397	54.8843
0.1348	8.0	648	1.2829	29.3764	54.5343	0.5451	55.6008
0.0884	9.0	729	1.2720	31.1738	52.4510	0.5651	56.4094
0.0774	10.0	810	1.2931	32.8888	52.2059	0.5692	57.6071
0.0545	11.0	891	1.3301	32.2780	53.1863	0.5715	56.1920
0.0387	12.0	972	1.3245	31.8118	53.4314	0.5602	56.7971
0.0288	13.0	1053	1.3220	33.1219	51.8382	0.5662	57.4929

josueu/m2m100-A100-es-zap-data

m2m100-A100-es-zap-data