635c9b8cacbe0d97df60ea14493c86d5

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

• Loss: 1.9852
• Data Size: 1.0
• Epoch Runtime: 54.5801
• Bleu: 3.3496

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	3.1696	0	4.1115	0.5260
No log	1	91	3.0036	0.0078	4.6242	0.5757
No log	2	182	2.8249	0.0156	8.2996	0.6023
No log	3	273	2.7593	0.0312	12.5513	0.6844
No log	4	364	2.6801	0.0625	19.8430	0.7439
No log	5	455	2.5680	0.125	25.2748	1.6307
No log	6	546	2.4544	0.25	29.6768	1.6493
0.2599	7	637	2.2993	0.5	40.8758	1.9173
2.3343	8.0	728	2.1336	1.0	67.9301	2.4068
2.1	9.0	819	2.0424	1.0	54.1273	2.7113
1.9163	10.0	910	1.9749	1.0	57.5055	2.9403
1.7507	11.0	1001	1.9348	1.0	61.0555	3.0826
1.622	12.0	1092	1.9196	1.0	56.5270	3.2027
1.4917	13.0	1183	1.9045	1.0	59.9145	3.1881
1.3648	14.0	1274	1.9053	1.0	58.2257	3.1194
1.245	15.0	1365	1.9187	1.0	57.7393	3.2542
1.1427	16.0	1456	1.9854	1.0	61.6901	3.3665
1.0307	17.0	1547	1.9852	1.0	54.5801	3.3496

Framework versions

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