distilbert

bsharpataz/distilbert

distilbert is a token classification model from bsharpataz. Use it when you need labels on individual words, such as names. 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

20% of all-time downloads

All-time downloads

Public

Repo size

522 MB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin261 MB · 100%

At a glance

Task: Token Classification
Library: transformers
License: apache-2.0
Model type: distilbert
Access: Public
Created: Nov 14, 2023
Updated: Nov 14, 2023
SHA: b337288e

Try a prompt

Base models

Model card

From the Hugging Face model README

distilbert

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

Loss: 0.0005
o precision: 0.9946
o recall: 0.9960
o f1: 0.9953
i precision: 0.9994
i recall: 0.9993
i f1: 0.9994
Weighted avg f1: 0.9989

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: 16
eval_batch_size: 16
seed: 426
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	`o` precision	`o` recall	`o` f1	`i` precision	`i` recall	`i` f1	Weighted avg f1
0.0187	0.08	500	0.0033	0.9363	0.9990	0.9666	0.9999	0.9906	0.9952	0.9918
0.002	0.16	1000	0.0014	0.9762	0.9947	0.9854	0.9993	0.9967	0.9980	0.9964
0.0016	0.24	1500	0.0012	0.9813	0.9918	0.9865	0.9989	0.9974	0.9981	0.9967
0.0015	0.32	2000	0.0012	0.9801	0.9960	0.9880	0.9994	0.9972	0.9983	0.9971
0.0013	0.4	2500	0.0010	0.9834	0.9960	0.9896	0.9994	0.9977	0.9986	0.9975
0.001	0.48	3000	0.0008	0.9881	0.9959	0.9920	0.9994	0.9983	0.9989	0.9980
0.0009	0.56	3500	0.0009	0.9854	0.9955	0.9904	0.9994	0.9980	0.9987	0.9977
0.0009	0.64	4000	0.0008	0.9883	0.9946	0.9914	0.9993	0.9984	0.9988	0.9979
0.0009	0.72	4500	0.0009	0.9935	0.9884	0.9910	0.9984	0.9991	0.9988	0.9978
0.0008	0.8	5000	0.0008	0.9913	0.9926	0.9920	0.9990	0.9988	0.9989	0.9981
0.0008	0.88	5500	0.0007	0.9874	0.9976	0.9925	0.9997	0.9982	0.9990	0.9982
0.0008	0.96	6000	0.0008	0.9924	0.9923	0.9923	0.9989	0.9990	0.9989	0.9982
0.0005	1.04	6500	0.0007	0.9924	0.9948	0.9936	0.9993	0.9990	0.9991	0.9985
0.0005	1.12	7000	0.0007	0.9885	0.9973	0.9929	0.9996	0.9984	0.9990	0.9983
0.0005	1.2	7500	0.0007	0.9890	0.9970	0.9930	0.9996	0.9985	0.9990	0.9983
0.0006	1.28	8000	0.0006	0.9927	0.9965	0.9946	0.9995	0.9990	0.9993	0.9987
0.0004	1.36	8500	0.0005	0.9934	0.9962	0.9948	0.9995	0.9991	0.9993	0.9987
0.0004	1.44	9000	0.0006	0.9941	0.9953	0.9947	0.9994	0.9992	0.9993	0.9987
0.0004	1.52	9500	0.0005	0.9940	0.9951	0.9946	0.9993	0.9992	0.9993	0.9987
0.0004	1.6	10000	0.0005	0.9942	0.9958	0.9950	0.9994	0.9992	0.9993	0.9988
0.0003	1.68	10500	0.0006	0.9940	0.9951	0.9945	0.9993	0.9992	0.9992	0.9987
0.0005	1.76	11000	0.0005	0.9953	0.9947	0.9950	0.9993	0.9994	0.9993	0.9988
0.0004	1.84	11500	0.0005	0.9944	0.9958	0.9951	0.9994	0.9992	0.9993	0.9988
0.0004	1.92	12000	0.0005	0.9943	0.9962	0.9953	0.9995	0.9992	0.9993	0.9989
0.0004	2.0	12500	0.0005	0.9946	0.9960	0.9953	0.9994	0.9993	0.9994	0.9989

Framework versions

Transformers 4.34.1
Pytorch 2.1.0
Datasets 2.14.6
Tokenizers 0.14.1

Task: Token Classification
Library: transformers
Type: distilbert
License: apache-2.0
Created: Nov 14, 2023
Updated: Nov 14, 2023