fine_tune_bert-base-cased

Chessmen/fine_tune_bert-base-cased

fine_tune_bert-base-cased is a question answering model from Chessmen. Use it when the input is a question plus a passage. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

28% of all-time downloads

All-time downloads

Public

Parameters

108M

8.6 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors431 MB · 100%

At a glance

Task: Question Answering
Library: transformers
License: apache-2.0
Model type: bert
Access: Public
Created: Sep 3, 2024
Updated: Sep 9, 2024
SHA: 3971161f

Try a prompt

Base models

Model card

From the Hugging Face model README

fine_tune_bert-base-cased

This model is a fine-tuned version of bert-base-cased on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.0842
Precision: 0.9376
Recall: 0.9541
F1: 0.9458
Accuracy: 0.9866

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: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.2667	1.0	220	0.0739	0.8619	0.9118	0.8862	0.9776
0.0602	2.0	440	0.0641	0.9109	0.9357	0.9231	0.9830
0.0361	3.0	660	0.0594	0.9187	0.9401	0.9293	0.9845
0.0234	4.0	880	0.0564	0.9233	0.9461	0.9346	0.9854
0.0164	5.0	1100	0.0585	0.9211	0.9465	0.9336	0.9856
0.0123	6.0	1320	0.0656	0.9212	0.9483	0.9346	0.9850
0.0084	7.0	1540	0.0639	0.9290	0.9514	0.9401	0.9864
0.0072	8.0	1760	0.0735	0.9325	0.9482	0.9403	0.9862
0.0051	9.0	1980	0.0745	0.9319	0.9488	0.9403	0.9856
0.0042	10.0	2200	0.0783	0.9308	0.9490	0.9398	0.9858
0.0034	11.0	2420	0.0782	0.9337	0.9509	0.9422	0.9862
0.0026	12.0	2640	0.0822	0.9328	0.9505	0.9416	0.9858
0.0019	13.0	2860	0.0785	0.9335	0.9525	0.9429	0.9862
0.0018	14.0	3080	0.0819	0.9382	0.9525	0.9453	0.9865
0.0015	15.0	3300	0.0846	0.9349	0.9524	0.9436	0.9863
0.0013	16.0	3520	0.0880	0.9353	0.9519	0.9435	0.9860
0.0012	17.0	3740	0.0846	0.9362	0.9527	0.9444	0.9864
0.001	18.0	3960	0.0868	0.9374	0.9532	0.9453	0.9864
0.0009	19.0	4180	0.0842	0.9381	0.9536	0.9458	0.9868
0.0009	20.0	4400	0.0842	0.9376	0.9541	0.9458	0.9866

Framework versions

Transformers 4.44.2
Pytorch 2.2.0+cu121
Datasets 2.21.0
Tokenizers 0.19.1

Task: Question Answering
Library: transformers
Type: bert
License: apache-2.0
Created: Sep 3, 2024
Updated: Sep 9, 2024