m2m100_418M-ja

This model is a fine-tuned version of facebook/m2m100_418M on the kde4 dataset. It achieves the following results on the evaluation set:

• Loss: nan
• Bleu: 0.0

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.002
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu
0.0	1.0	59084	nan	0.0
0.0	2.0	118168	nan	0.0
0.0	3.0	177252	nan	0.0
0.0	4.0	236336	nan	0.0
0.0	5.0	295420	nan	0.0
0.0	6.0	354504	nan	0.0
0.0	7.0	413588	nan	0.0
0.0	8.0	472672	nan	0.0
0.0	9.0	531756	nan	0.0
0.0	10.0	590840	nan	0.0
0.0	11.0	649924	nan	0.0
0.0	12.0	709008	nan	0.0
0.0	13.0	768092	nan	0.0
0.0	14.0	827176	nan	0.0
0.0	15.0	886260	nan	0.0
0.0	16.0	945344	nan	0.0
0.0	17.0	1004428	nan	0.0
0.0	18.0	1063512	nan	0.0
0.0	19.0	1122596	nan	0.0
0.0	20.0	1181680	nan	0.0
0.0	21.0	1240764	nan	0.0
0.0	22.0	1299848	nan	0.0
0.0	23.0	1358932	nan	0.0
0.0	24.0	1418016	nan	0.0
0.0	25.0	1477100	nan	0.0
0.0	26.0	1536184	nan	0.0
0.0	27.0	1595268	nan	0.0
0.0	28.0	1654352	nan	0.0
0.0	29.0	1713436	nan	0.0
0.0	30.0	1772520	nan	0.0
0.0	31.0	1831604	nan	0.0
0.0	32.0	1890688	nan	0.0
0.0	33.0	1949772	nan	0.0
0.0	34.0	2008856	nan	0.0
0.0	35.0	2067940	nan	0.0
0.0	36.0	2127024	nan	0.0
0.0	37.0	2186108	nan	0.0
0.0	38.0	2245192	nan	0.0
0.0	39.0	2304276	nan	0.0
0.0	40.0	2363360	nan	0.0
0.0	41.0	2422444	nan	0.0
0.0	42.0	2481528	nan	0.0
0.0	43.0	2540612	nan	0.0
0.0	44.0	2599696	nan	0.0
0.0	45.0	2658780	nan	0.0
0.0	46.0	2717864	nan	0.0
0.0	47.0	2776948	nan	0.0
0.0	48.0	2836032	nan	0.0
0.0	49.0	2895116	nan	0.0
0.0	50.0	2954200	nan	0.0

Framework versions

• Transformers 4.27.4
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.2