0d888938075e1f2c99ed446e64e2b0c3

This model is a fine-tuned version of google/umt5-xl on the Helsinki-NLP/opus_books [fr-nl] dataset. It achieves the following results on the evaluation set:

• Loss: 1.4333
• Data Size: 1.0
• Epoch Runtime: 505.8870
• Bleu: 13.4520

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	5.5362	0	34.8559	1.8596
No log	1	1000	3.6002	0.0078	38.6913	9.5468
No log	2	2000	2.7465	0.0156	47.3274	15.4732
No log	3	3000	2.2396	0.0312	57.5305	19.6151
0.1131	4	4000	1.9828	0.0625	74.1762	8.1750
2.2443	5	5000	1.7259	0.125	102.1856	9.3906
0.1237	6	6000	1.6146	0.25	158.8772	10.5212
0.1571	7	7000	1.5000	0.5	272.0392	11.4013
1.5502	8.0	8000	1.3952	1.0	516.9643	12.3219
1.3822	9.0	9000	1.3633	1.0	519.2476	12.8708
1.2293	10.0	10000	1.3496	1.0	514.9750	13.1942
1.1165	11.0	11000	1.3517	1.0	502.6161	13.3377
0.9984	12.0	12000	1.3675	1.0	505.7465	13.4797
0.9015	13.0	13000	1.3929	1.0	501.8769	13.4853
0.8082	14.0	14000	1.4333	1.0	505.8870	13.4520

Framework versions

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