e5db469994486b3046d3bee20f5b59b2

This model is a fine-tuned version of google/mt5-xl on the Helsinki-NLP/opus_books [es-ru] dataset. It achieves the following results on the evaluation set:

• Loss: 1.6967
• Data Size: 1.0
• Epoch Runtime: 202.8000
• Bleu: 7.5478

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	8.4261	0	13.4485	0.0172
No log	1	419	4.1196	0.0078	15.1340	0.1221
No log	2	838	3.2758	0.0156	20.7725	0.3325
0.1383	3	1257	2.5093	0.0312	27.2346	0.6518
0.1383	4	1676	1.9884	0.0625	32.5211	4.2887
0.1635	5	2095	1.8328	0.125	46.3963	4.9586
0.2874	6	2514	1.7391	0.25	66.7812	5.4252
1.9274	7	2933	1.6624	0.5	114.3662	6.4305
1.7325	8.0	3352	1.5978	1.0	202.9380	6.9353
1.5661	9.0	3771	1.5676	1.0	198.7221	7.4073
1.4193	10.0	4190	1.5643	1.0	201.4874	7.5302
1.2444	11.0	4609	1.5840	1.0	200.1250	7.6954
1.1652	12.0	5028	1.6043	1.0	200.3730	7.7712
1.0323	13.0	5447	1.6456	1.0	201.1470	7.5048
0.9682	14.0	5866	1.6967	1.0	202.8000	7.5478

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1