cd267d808df29a76255579024ed94db0

This model is a fine-tuned version of google-t5/t5-base on the Helsinki-NLP/opus_books [fr-pt] dataset. It achieves the following results on the evaluation set:

• Loss: 1.8689
• Data Size: 1.0
• Epoch Runtime: 12.7755
• Bleu: 5.7285

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	3.7252	0	1.3765	0.5767
No log	1	31	3.6489	0.0078	1.7274	0.5662
No log	2	62	3.4703	0.0156	1.9435	0.6171
No log	3	93	3.2871	0.0312	2.3366	0.7242
No log	4	124	3.1396	0.0625	2.5323	0.8164
No log	5	155	2.9223	0.125	2.8502	1.4783
No log	6	186	2.6968	0.25	3.9540	1.4467
0.5063	7	217	2.4922	0.5	6.4702	2.5828
0.5063	8.0	248	2.3385	1.0	10.0464	2.5829
1.906	9.0	279	2.2540	1.0	9.6595	2.8171
2.5147	10.0	310	2.1872	1.0	9.4018	3.3787
2.5147	11.0	341	2.1315	1.0	11.9825	3.5915
2.3771	12.0	372	2.0891	1.0	10.8245	3.9220
2.2552	13.0	403	2.0584	1.0	10.7654	4.1933
2.2552	14.0	434	2.0289	1.0	12.6130	4.2683
2.1522	15.0	465	2.0015	1.0	12.9307	4.4838
2.1522	16.0	496	1.9792	1.0	11.8939	4.9288
2.06	17.0	527	1.9677	1.0	10.7586	5.0028
1.9767	18.0	558	1.9496	1.0	11.1887	5.1620
1.9767	19.0	589	1.9373	1.0	10.3486	5.1661
1.9061	20.0	620	1.9266	1.0	10.3754	5.1572
1.8546	21.0	651	1.9186	1.0	10.8199	5.2165
1.8546	22.0	682	1.9068	1.0	10.5035	5.3257
1.7996	23.0	713	1.8994	1.0	10.8727	5.3570
1.7996	24.0	744	1.8960	1.0	11.5351	5.2844
1.7327	25.0	775	1.8832	1.0	12.7932	5.3954
1.6918	26.0	806	1.8809	1.0	12.9421	5.3482
1.6918	27.0	837	1.8776	1.0	10.9155	5.4996
1.6355	28.0	868	1.8725	1.0	10.6607	5.5058
1.6355	29.0	899	1.8741	1.0	11.8382	5.5790
1.5971	30.0	930	1.8680	1.0	11.1133	5.5874
1.5452	31.0	961	1.8781	1.0	11.4915	5.6126
1.5452	32.0	992	1.8733	1.0	11.6596	5.7994
1.508	33.0	1023	1.8658	1.0	11.8230	5.5793
1.4665	34.0	1054	1.8665	1.0	13.0364	5.7622
1.4665	35.0	1085	1.8705	1.0	11.7001	5.6563
1.4258	36.0	1116	1.8670	1.0	12.3637	5.7373
1.4258	37.0	1147	1.8689	1.0	12.7755	5.7285

Framework versions

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