4dbd9f3dfeab5aaabcbe645497570b42

This model is a fine-tuned version of google/mt5-large on the Helsinki-NLP/opus_books [fr-no] dataset. It achieves the following results on the evaluation set:

• Loss: 2.1591
• Data Size: 1.0
• Epoch Runtime: 41.6813
• Bleu: 5.9160

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	21.8402	0	3.6309	0.0053
No log	1	86	23.5311	0.0078	4.8345	0.0051
No log	2	172	27.4261	0.0156	6.7790	0.0029
No log	3	258	27.9904	0.0312	8.9397	0.0025
No log	4	344	26.7916	0.0625	12.6691	0.0019
1.396	5	430	21.7076	0.125	14.7820	0.0018
6.0952	6	516	16.4132	0.25	17.6422	0.0027
6.6518	7	602	8.5601	0.5	26.3368	0.0033
5.867	8.0	688	3.5917	1.0	44.1223	0.1667
3.5936	9.0	774	2.4151	1.0	40.4858	3.6838
2.8462	10.0	860	2.2648	1.0	40.5889	4.2880
2.6527	11.0	946	2.2103	1.0	41.6765	4.8279
2.4108	12.0	1032	2.1808	1.0	42.3148	5.0905
2.3374	13.0	1118	2.1662	1.0	41.2657	5.2547
2.209	14.0	1204	2.1465	1.0	40.5995	5.3263
2.1111	15.0	1290	2.1578	1.0	41.1459	5.5766
2.0158	16.0	1376	2.1539	1.0	41.6166	5.6148
1.9092	17.0	1462	2.1503	1.0	42.8685	5.6329
1.8442	18.0	1548	2.1591	1.0	41.6813	5.9160

Framework versions

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