Medical-NER-finetuned-ner

NeverLearn/Medical-NER-finetuned-ner

Medical-NER-finetuned-ner is a token classification model from NeverLearn. Use it when you need labels on individual words, such as names. It is set up for transformers. The card lists the license as mit.

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

Downloads · 30 days

2% of all-time downloads

All-time downloads

1.3K

Public

Parameters

184M

10.3 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors736 MB · 99%

At a glance

Task: Token Classification
Library: transformers
License: mit
Model type: deberta-v2
Access: Public
Created: Apr 5, 2024
Updated: Apr 15, 2024
SHA: be451be3

Try a prompt

Base models

Datasets

maccrobat_biomedical_ner

Model card

From the Hugging Face model README

Medical-NER-finetuned-ner

This model is a fine-tuned version of Clinical-AI-Apollo/Medical-NER on the maccrobat_biomedical_ner dataset. It achieves the following results on the evaluation set:

Loss: 0.5635
Precision: 0.8425
Recall: 0.8538
F1: 0.8481
Accuracy: 0.9046

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: 8.26814930103799e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	20	0.3925	0.8364	0.8307	0.8335	0.8912
No log	2.0	40	0.3671	0.8266	0.8529	0.8395	0.8954
No log	3.0	60	0.4077	0.8073	0.8388	0.8227	0.8843
No log	4.0	80	0.3630	0.8531	0.8463	0.8497	0.9045
No log	5.0	100	0.3717	0.8413	0.8484	0.8449	0.9017
No log	6.0	120	0.3721	0.8433	0.8425	0.8429	0.9015
No log	7.0	140	0.3679	0.8553	0.8529	0.8541	0.9069
No log	8.0	160	0.3840	0.8394	0.8504	0.8449	0.9012
No log	9.0	180	0.4124	0.8430	0.8520	0.8475	0.9040
No log	10.0	200	0.4328	0.8358	0.8450	0.8404	0.9004
No log	11.0	220	0.4395	0.8395	0.8552	0.8473	0.9033
No log	12.0	240	0.4490	0.8399	0.8490	0.8444	0.9011
No log	13.0	260	0.4592	0.8411	0.8497	0.8454	0.9027
No log	14.0	280	0.4623	0.8435	0.8525	0.8480	0.9047
No log	15.0	300	0.4858	0.8416	0.8540	0.8478	0.9040
No log	16.0	320	0.4986	0.8393	0.8499	0.8446	0.9019
No log	17.0	340	0.5152	0.8367	0.8474	0.8420	0.9012
No log	18.0	360	0.5138	0.8474	0.8508	0.8491	0.9055
No log	19.0	380	0.5414	0.8384	0.8488	0.8436	0.9015
No log	20.0	400	0.5483	0.8401	0.8508	0.8454	0.9029
No log	21.0	420	0.5465	0.8386	0.8454	0.8420	0.9008
No log	22.0	440	0.5463	0.8410	0.8520	0.8465	0.9034
No log	23.0	460	0.5434	0.8441	0.8545	0.8493	0.9053
No log	24.0	480	0.5516	0.8439	0.8493	0.8466	0.9041
0.1398	25.0	500	0.5618	0.8398	0.8518	0.8458	0.9032
0.1398	26.0	520	0.5583	0.8428	0.8550	0.8489	0.9046
0.1398	27.0	540	0.5632	0.8427	0.8524	0.8475	0.9042
0.1398	28.0	560	0.5674	0.8393	0.8522	0.8457	0.9029
0.1398	29.0	580	0.5625	0.8429	0.8527	0.8478	0.9046
0.1398	30.0	600	0.5635	0.8425	0.8538	0.8481	0.9046

Framework versions

Transformers 4.39.3
Pytorch 2.2.1
Datasets 2.18.0
Tokenizers 0.15.2

Task: Token Classification
Library: transformers
Type: deberta-v2
License: mit
Created: Apr 5, 2024
Updated: Apr 15, 2024