You can deploy this model and start earning money today!

contemmcm/
7add9fb0ab60519287384ec4d3e5c21a

Name: 7add9fb0ab60519287384ec4d3e5c21a
Rating: 2.5 (8 reviews)

generated_from_trainer

Views

No views yet

7add9fb0ab60519287384ec4d3e5c21a

This model is a fine-tuned version of google/umt5-xl on the Helsinki-NLP/opus_books [de-ru] dataset. It achieves the following results on the evaluation set:

Loss: 1.3894
Data Size: 1.0
Epoch Runtime: 237.4442
Bleu: 11.7682

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	5.5143	0	15.6591	2.2168
No log	1	434	3.9378	0.0078	18.0215	6.9689
No log	2	868	3.1265	0.0156	24.5175	12.2193
No log	3	1302	2.4663	0.0312	34.3183	18.1388
No log	4	1736	1.9607	0.0625	41.9881	22.1906
0.1281	5	2170	1.6461	0.125	55.7824	8.4147
1.8911	6	2604	1.4763	0.25	80.2688	9.1950
1.6728	7	3038	1.3879	0.5	131.1561	9.9621
1.4772	8.0	3472	1.3040	1.0	237.1828	10.7771
1.2574	9.0	3906	1.2655	1.0	236.6050	11.1237
1.1141	10.0	4340	1.2649	1.0	234.8766	11.3560
0.9724	11.0	4774	1.2726	1.0	238.8724	11.5951
0.8455	12.0	5208	1.3096	1.0	239.6048	11.6940
0.7614	13.0	5642	1.3370	1.0	240.2006	11.7613
0.6633	14.0	6076	1.3894	1.0	237.4442	11.7682

Framework versions

Transformers 4.57.0
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.1

contemmcm/7add9fb0ab60519287384ec4d3e5c21a

7add9fb0ab60519287384ec4d3e5c21a

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

contemmcm/
7add9fb0ab60519287384ec4d3e5c21a