f4a8fa8d43c5864ae147b1bf5058e7d2

This model is a fine-tuned version of google-t5/t5-3b on the Helsinki-NLP/opus_books [fr-ru] dataset. It achieves the following results on the evaluation set:

• Loss: 1.2355
• Data Size: 1.0
• Epoch Runtime: 99.1010
• Bleu: 14.5219

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	2.0371	0	6.3108	0.4786
No log	1	204	1.8244	0.0078	7.6813	1.3987
No log	2	408	1.7120	0.0156	11.4930	2.3632
No log	3	612	1.6321	0.0312	17.1863	2.6770
No log	4	816	1.5794	0.0625	21.1195	3.1989
No log	5	1020	1.4983	0.125	25.7803	4.2789
0.1321	6	1224	1.4142	0.25	39.3120	5.9296
1.4497	7	1428	1.3248	0.5	59.4138	7.5477
1.3136	8.0	1632	1.2217	1.0	98.8663	9.7442
1.1725	9.0	1836	1.1721	1.0	95.4312	11.1821
1.0815	10.0	2040	1.1393	1.0	94.1836	11.9105
0.9702	11.0	2244	1.1226	1.0	95.6183	12.6533
0.8763	12.0	2448	1.1347	1.0	95.0917	13.4398
0.8101	13.0	2652	1.1217	1.0	92.7829	13.5905
0.7315	14.0	2856	1.1357	1.0	93.8128	14.3816
0.6768	15.0	3060	1.1624	1.0	93.9693	14.4867
0.5977	16.0	3264	1.2080	1.0	95.4702	14.8598
0.5563	17.0	3468	1.2355	1.0	99.1010	14.5219

Framework versions

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