Edit model card

zhko_mbartLarge_19p_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: 1.4881
  • Bleu: 9.8019
  • Gen Len: 14.5611

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: 1500
  • num_epochs: 15

Training results

Training Loss Epoch Step Validation Loss Bleu Gen Len
1.1911 1.79 5000 1.4881 9.7977 14.6128
0.6536 3.59 10000 1.6061 13.3897 14.9179
0.3665 5.38 15000 1.7929 14.0018 15.2051
0.194 7.18 20000 1.9399 14.7102 14.7308
0.1004 8.97 25000 2.0679 14.9684 14.8811

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.1.0+cu121
  • Datasets 2.14.6
  • Tokenizers 0.14.1
Downloads last month
1

Finetuned from