36e4a4f4479205827cbaeb7182497736

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

• Loss: 2.0411
• Data Size: 1.0
• Epoch Runtime: 37.1142
• Bleu: 7.5054

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	24.0301	0	3.2338	0.0077
No log	1	74	22.3734	0.0078	3.4871	0.0069
No log	2	148	18.0931	0.0156	6.8167	0.0107
0.8365	3	222	17.6183	0.0312	8.9247	0.0115
0.8365	4	296	15.7380	0.0625	11.1980	0.0115
1.4537	5	370	16.1153	0.125	14.3266	0.0080
1.4537	6	444	13.7049	0.25	18.2127	0.0068
3.7567	7	518	4.4408	0.5	24.1257	0.0363
4.4368	8.0	592	2.5668	1.0	38.0884	0.7065
3.1454	9.0	666	2.2962	1.0	35.8996	11.8106
2.8331	10.0	740	2.1845	1.0	35.5927	4.8443
2.5558	11.0	814	2.1278	1.0	37.6358	5.0953
2.4679	12.0	888	2.0948	1.0	38.5283	5.5664
2.2716	13.0	962	2.0673	1.0	35.9105	6.3039
2.2181	14.0	1036	2.0499	1.0	37.5357	7.0011
2.0843	15.0	1110	2.0392	1.0	35.0586	6.9639
2.0004	16.0	1184	2.0340	1.0	36.1249	7.2430
1.9004	17.0	1258	2.0303	1.0	35.3188	7.2581
1.8689	18.0	1332	2.0420	1.0	36.0142	7.0671
1.7816	19.0	1406	2.0426	1.0	34.6889	7.1300
1.7352	20.0	1480	2.0407	1.0	36.7340	7.3224
1.65	21.0	1554	2.0411	1.0	37.1142	7.5054

Framework versions

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