73bb82a1389478d70fce71a84ce2ce75

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

• Loss: 1.8587
• Data Size: 1.0
• Epoch Runtime: 57.7931
• Bleu: 2.5157

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	3.3495	0	3.9895	0.0937
No log	1	74	2.9422	0.0078	4.9811	0.1676
No log	2	148	2.6403	0.0156	10.4416	0.4590
0.1177	3	222	2.5376	0.0312	14.7265	0.9074
0.1177	4	296	2.4607	0.0625	21.4291	0.9322
0.2048	5	370	2.3894	0.125	26.0582	0.9688
0.2048	6	444	2.3037	0.25	28.8315	0.8164
0.5766	7	518	2.2143	0.5	40.0099	0.9857
1.6131	8.0	592	2.0921	1.0	65.7096	1.1593
2.2401	9.0	666	2.0161	1.0	56.2678	1.3410
2.1418	10.0	740	1.9588	1.0	59.8908	1.5716
1.9774	11.0	814	1.9149	1.0	57.1123	1.6924
1.8915	12.0	888	1.8871	1.0	59.7441	1.7405
1.7544	13.0	962	1.8661	1.0	56.6976	1.9092
1.6864	14.0	1036	1.8484	1.0	57.0622	2.0409
1.5622	15.0	1110	1.8233	1.0	58.9231	2.1094
1.4977	16.0	1184	1.8393	1.0	59.3126	2.2520
1.401	17.0	1258	1.8257	1.0	61.1558	2.4829
1.3292	18.0	1332	1.8523	1.0	56.2815	2.5341
1.2325	19.0	1406	1.8587	1.0	57.7931	2.5157

Framework versions

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