964a2f79c203348bba76d60ca938cd2d

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

• Loss: 1.8123
• Data Size: 1.0
• Epoch Runtime: 43.0591
• Bleu: 6.1264

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	22.2111	0	3.7215	0.0208
No log	1	88	21.6220	0.0078	4.5621	0.0193
No log	2	176	16.0156	0.0156	5.8968	0.0190
No log	3	264	11.5975	0.0312	7.7077	0.0281
No log	4	352	6.8749	0.0625	10.3325	0.0261
No log	5	440	5.7621	0.125	13.3582	0.0484
0.9257	6	528	3.2954	0.25	17.3148	0.1578
1.5513	7	616	2.5954	0.5	27.3271	0.4223
2.923	8.0	704	2.0603	1.0	43.4494	9.3587
2.6132	9.0	792	1.9396	1.0	42.2230	3.6443
2.3564	10.0	880	1.8772	1.0	41.0739	4.0174
2.1864	11.0	968	1.8413	1.0	42.9801	4.4609
2.0437	12.0	1056	1.8230	1.0	41.3114	5.4799
1.9651	13.0	1144	1.8168	1.0	42.6002	5.4061
1.8622	14.0	1232	1.8016	1.0	40.8660	5.8221
1.7951	15.0	1320	1.8066	1.0	43.6785	5.9283
1.7245	16.0	1408	1.8128	1.0	43.5209	6.1232
1.6505	17.0	1496	1.8127	1.0	42.7977	6.1399
1.5791	18.0	1584	1.8123	1.0	43.0591	6.1264

Framework versions

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