You can deploy this model and start earning money today!

contemmcm/
ec410788ec07b2840be0aab0d6cb7f79

generated_from_trainer

facebook/mbart-large-50-many-to-one-mmt

finetune

endpoints_compatible

Views

No views yet

ec410788ec07b2840be0aab0d6cb7f79

This model is a fine-tuned version of facebook/mbart-large-50-many-to-one-mmt on the Helsinki-NLP/opus_books [fr-no] dataset. It achieves the following results on the evaluation set:

Loss: 4.2691
Data Size: 1.0
Epoch Runtime: 24.6138
Bleu: 5.8363

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: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 32
total_eval_batch_size: 32
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: constant
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Bleu
No log	0	0	8.6901	0	2.3714	0.2792
No log	1	86	7.9900	0.0078	3.3065	0.3097
No log	2	172	7.3854	0.0156	3.9775	0.3983
No log	3	258	6.9651	0.0312	4.8523	0.4699
No log	4	344	6.1491	0.0625	6.3021	0.6245
0.3483	5	430	5.5111	0.125	8.4834	1.0717
1.2978	6	516	4.8931	0.25	11.0534	1.6614
1.6334	7	602	4.3708	0.5	14.8638	2.2289
2.2536	8.0	688	3.9487	1.0	25.6906	2.9909
3.1116	9.0	774	3.7822	1.0	26.6359	5.1185
2.5527	10.0	860	3.7889	1.0	24.7960	4.4563
2.1262	11.0	946	3.9449	1.0	25.5257	8.2390
1.5837	12.0	1032	4.1394	1.0	26.0984	7.1303
1.2872	13.0	1118	4.2691	1.0	24.6138	5.8363

Framework versions

Transformers 4.57.0
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.1

contemmcm/ec410788ec07b2840be0aab0d6cb7f79

ec410788ec07b2840be0aab0d6cb7f79

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

contemmcm/
ec410788ec07b2840be0aab0d6cb7f79