jako_mbartLarge_13p_run1

This model is a fine-tuned version of facebook/mbart-large-50-many-to-many-mmt on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0819
• Bleu: 29.1055
• Gen Len: 18.2731

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: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• total_eval_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 300
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
1.3826	0.39	1500	1.2989	22.0729	18.9717
1.1964	0.78	3000	1.1630	25.4863	19.1908
0.9449	1.17	4500	1.1125	27.385	18.2955
0.8102	1.56	6000	1.0920	28.0041	18.6572
0.7692	1.95	7500	1.0819	29.1055	18.2731
0.5741	2.34	9000	1.1369	28.1574	18.3485
0.5198	2.73	10500	1.1538	28.657	18.4527
0.4532	3.12	12000	1.1582	28.6914	18.4562
0.3466	3.51	13500	1.2048	28.8955	18.427

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu121
• Datasets 2.14.6
• Tokenizers 0.14.1