400c5511b6dea2d366b982d82c7f2f47

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

• Loss: 1.7447
• Data Size: 1.0
• Epoch Runtime: 89.7493
• Bleu: 8.7516

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	4.5136	0	7.0465	0.4966
No log	1	367	4.1107	0.0078	8.0846	0.7376
No log	2	734	3.5075	0.0156	8.7588	0.9149
No log	3	1101	3.1627	0.0312	10.9472	1.1631
No log	4	1468	3.0039	0.0625	12.4164	1.5942
0.1766	5	1835	2.8927	0.125	17.6485	2.0877
3.0996	6	2202	2.7307	0.25	27.5012	2.5708
2.8871	7	2569	2.5677	0.5	47.1074	4.1272
2.6588	8.0	2936	2.3864	1.0	92.7098	5.1262
2.4844	9.0	3303	2.2674	1.0	92.5091	5.4158
2.3644	10.0	3670	2.1871	1.0	88.8597	5.8018
2.2802	11.0	4037	2.1226	1.0	87.4977	6.1405
2.2119	12.0	4404	2.0682	1.0	88.0479	6.4334
2.1701	13.0	4771	2.0277	1.0	85.0628	6.5505
2.077	14.0	5138	1.9844	1.0	87.9358	6.8263
2.0346	15.0	5505	1.9514	1.0	86.9127	7.0307
1.9813	16.0	5872	1.9358	1.0	86.8131	7.2225
1.9399	17.0	6239	1.9115	1.0	87.6552	7.4084
1.8997	18.0	6606	1.8905	1.0	91.3228	7.6722
1.8824	19.0	6973	1.8781	1.0	90.9620	7.5620
1.8423	20.0	7340	1.8488	1.0	88.2698	7.7587
1.7856	21.0	7707	1.8354	1.0	88.6626	7.8544
1.7663	22.0	8074	1.8322	1.0	89.3511	7.7950
1.7213	23.0	8441	1.8203	1.0	87.7341	7.9426
1.6967	24.0	8808	1.8010	1.0	90.4521	8.0048
1.6952	25.0	9175	1.7929	1.0	91.0979	7.9804
1.6379	26.0	9542	1.7913	1.0	91.5110	7.8794
1.6116	27.0	9909	1.7820	1.0	86.8504	8.0096
1.583	28.0	10276	1.7778	1.0	85.1309	8.2697
1.5875	29.0	10643	1.7725	1.0	91.7085	8.2960
1.5704	30.0	11010	1.7684	1.0	92.7509	8.1364
1.5249	31.0	11377	1.7576	1.0	90.7023	8.2841
1.5196	32.0	11744	1.7593	1.0	94.4428	8.3993
1.471	33.0	12111	1.7525	1.0	91.7521	8.3661
1.4632	34.0	12478	1.7573	1.0	90.2000	8.3085
1.4363	35.0	12845	1.7428	1.0	89.6053	8.6387
1.4309	36.0	13212	1.7485	1.0	95.4347	8.8372
1.4165	37.0	13579	1.7513	1.0	84.3171	8.7722
1.4042	38.0	13946	1.7471	1.0	87.1024	8.8673
1.3828	39.0	14313	1.7447	1.0	89.7493	8.7516

Framework versions

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