metadata-cls-no-gov-8k-vnnic-xml

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3308
  • Accuracy: 0.9345
  • F1: 0.7072

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: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Accuracy F1
No log 0.5703 150 0.2151 0.9411 0.4945
0.5567 1.1407 300 0.2789 0.9121 0.4579
0.5567 1.7110 450 0.2676 0.9130 0.6930
0.2703 2.2814 600 0.1926 0.9345 0.7071
0.2703 2.8517 750 0.1984 0.9411 0.7390
0.2055 3.4221 900 0.2503 0.9205 0.6689
0.1573 3.9924 1050 0.1895 0.9476 0.7332
0.1573 4.5627 1200 0.3001 0.9130 0.6974
0.1287 5.1331 1350 0.1834 0.9495 0.7302
0.1287 5.7034 1500 0.2586 0.9429 0.7345
0.1022 6.2738 1650 0.3486 0.9261 0.7055
0.1022 6.8441 1800 0.3064 0.9317 0.7053
0.085 7.4144 1950 0.3445 0.9308 0.7086
0.0689 7.9848 2100 0.3342 0.9336 0.7130
0.0689 8.5551 2250 0.3272 0.9345 0.7074
0.0525 9.1255 2400 0.3391 0.9345 0.7030
0.0525 9.6958 2550 0.3308 0.9345 0.7072

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.1.2
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
23
Safetensors
Model size
278M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for gechim/metadata-cls-no-gov-8k-vnnic-xml

Finetuned
(2693)
this model