Edit model card

BERT_pretraining_h_100

This model is a fine-tuned version of bert-large-uncased on the gokuls/wiki_book_corpus_complete_processed_bert_dataset dataset. It achieves the following results on the evaluation set:

  • Loss: 7.2715
  • Accuracy: 0.0465

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: 0.0001
  • train_batch_size: 36
  • eval_batch_size: 36
  • seed: 10
  • distributed_type: multi-GPU
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100000
  • num_epochs: 100

Training results

Training Loss Epoch Step Validation Loss Accuracy
6.5136 0.06 10000 6.4841 0.1332
6.3225 0.12 20000 6.2865 0.1452
6.0806 0.18 30000 6.1945 0.1482
6.1449 0.25 40000 6.1284 0.1497
5.7325 0.31 50000 5.8403 0.1609
4.0177 0.37 60000 3.7789 0.3887
3.3942 0.43 70000 3.1742 0.4638
3.2801 0.49 80000 3.0618 0.4775
7.2562 0.55 90000 7.2798 0.0432
7.226 0.61 100000 7.2771 0.0465
7.2174 0.68 110000 7.2764 0.0465
7.232 0.74 120000 7.2745 0.0465
7.2003 0.8 130000 7.2730 0.0465
7.0964 0.86 140000 7.2725 0.0466
7.5174 0.92 150000 7.2729 0.0465
7.2674 0.98 160000 7.2729 0.0465
7.2044 1.04 170000 7.2729 0.0466

Framework versions

  • Transformers 4.37.1
  • Pytorch 2.1.2+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.1
Downloads last month
1
Safetensors
Model size
366M 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 gokuls/BERT_pretraining_h_100

Finetuned
(103)
this model

Dataset used to train gokuls/BERT_pretraining_h_100

Evaluation results

  • Accuracy on gokuls/wiki_book_corpus_complete_processed_bert_dataset
    self-reported
    0.047