fbbd17bd91d6f9404ef6a1c87c86debb

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

• Loss: 1.5335
• Data Size: 1.0
• Epoch Runtime: 32.1254
• Bleu: 9.6740

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	7.0196	0	1.7101	0.0147
No log	1	31	6.4985	0.0078	2.0311	0.0218
No log	2	62	5.0699	0.0156	7.0458	0.0241
No log	3	93	4.3467	0.0312	11.4814	0.0886
No log	4	124	3.3867	0.0625	16.9651	0.3731
No log	5	155	2.8183	0.125	23.2827	0.7309
No log	6	186	2.3754	0.25	25.3697	0.9775
0.6263	7	217	1.8119	0.5	26.8315	1.7286
0.6263	8.0	248	1.6395	1.0	31.8482	7.3810
1.5177	9.0	279	1.5286	1.0	37.3552	8.5945
1.7166	10.0	310	1.4919	1.0	29.2609	8.4978
1.7166	11.0	341	1.4923	1.0	31.7986	9.1875
1.4202	12.0	372	1.5036	1.0	37.8897	8.9715
1.1968	13.0	403	1.5168	1.0	28.6904	10.2552
1.1968	14.0	434	1.5335	1.0	32.1254	9.6740

Framework versions

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