README.md

metadata

license: mit
tags:
  - generated_from_trainer
datasets:
  - cartesinus/iva_mt_wslot
metrics:
  - bleu
model-index:
  - name: iva_mt_wslot-m2m100_418M-en-es
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: iva_mt_wslot
          type: iva_mt_wslot
          config: en-es
          split: validation
          args: en-es
        metrics:
          - name: Bleu
            type: bleu
            value: 69.2836
language:
  - en
  - es

iva_mt_wslot-m2m100_418M-en-es

This model is a fine-tuned version of facebook/m2m100_418M on the iva_mt_wslot dataset. It achieves the following results on the evaluation set:

• Loss: 0.0115
• Bleu: 69.2836
• Gen Len: 20.2064

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 7
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.0135	1.0	2104	0.0122	66.8284	20.2851
0.009	2.0	4208	0.0112	68.1164	20.1501
0.0067	3.0	6312	0.0110	68.256	20.0603
0.0051	4.0	8416	0.0110	68.7002	20.1219
0.0037	5.0	10520	0.0112	68.699	20.2733
0.0027	6.0	12624	0.0113	68.9916	20.209
0.0023	7.0	14728	0.0115	69.2836	20.2064

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.3