metadata

library_name: transformers
license: apache-2.0
base_model: google/bert_uncased_L-2_H-128_A-2
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
model-index:
  - name: bert_uncased_L-2_H-128_A-2_mrpc
    results: []

bert_uncased_L-2_H-128_A-2_mrpc

This model is a fine-tuned version of google/bert_uncased_L-2_H-128_A-2 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.5587
Accuracy: 0.7377
F1: 0.8283
Combined Score: 0.7830

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: 256
eval_batch_size: 256
seed: 10
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Combined Score
0.6493	1.0	15	0.6227	0.6838	0.8122	0.7480
0.6257	2.0	30	0.6134	0.6838	0.8122	0.7480
0.6126	3.0	45	0.6052	0.6838	0.8122	0.7480
0.6036	4.0	60	0.5954	0.6961	0.8176	0.7569
0.5897	5.0	75	0.5879	0.6985	0.8167	0.7576
0.5781	6.0	90	0.5741	0.7034	0.8158	0.7596
0.5635	7.0	105	0.5711	0.7108	0.8201	0.7655
0.5429	8.0	120	0.5674	0.7132	0.8208	0.7670
0.5228	9.0	135	0.5685	0.7206	0.8252	0.7729
0.5057	10.0	150	0.5497	0.7304	0.8281	0.7793
0.4856	11.0	165	0.5438	0.7377	0.8293	0.7835
0.4657	12.0	180	0.5352	0.7328	0.8233	0.7781
0.4447	13.0	195	0.5435	0.7402	0.8323	0.7862
0.4175	14.0	210	0.5562	0.7402	0.8328	0.7865
0.4039	15.0	225	0.5759	0.7426	0.8357	0.7892
0.3964	16.0	240	0.5610	0.7377	0.8299	0.7838
0.3735	17.0	255	0.5587	0.7377	0.8283	0.7830

Framework versions

Transformers 4.46.3
Pytorch 2.2.1+cu118
Datasets 2.17.0
Tokenizers 0.20.3