mpnet-base-airlines-news-multi-label

dahe827/mpnet-base-airlines-news-multi-label

mpnet-base-airlines-news-multi-label is a text classification model from dahe827. Use it when you need a label for a piece of text. It is set up for transformers.

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

Downloads · 30 days

2% of all-time downloads

All-time downloads

317

Public

Parameters

110M

1.3 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors439 MB · 100%

At a glance

Task: Text Classification
Library: transformers
Model type: mpnet
Access: Public
Created: May 27, 2024
Updated: Jun 5, 2024
SHA: fa881e97

Try a prompt

Base models

microsoft/mpnet-base

Model card

From the Hugging Face model README

mpnet-base-airlines-news-multi-label

This model is a fine-tuned version of microsoft/mpnet-base on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.2601
F1: 0.8921
Roc Auc: 0.6253

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: 7e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 65

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc
No log	1.0	57	0.3852	0.8161	0.5
No log	2.0	114	0.3612	0.8161	0.5
No log	3.0	171	0.3569	0.8161	0.5
No log	4.0	228	0.3515	0.8161	0.5
No log	5.0	285	0.3453	0.8161	0.5
No log	6.0	342	0.3403	0.8161	0.5
No log	7.0	399	0.3345	0.8161	0.5
No log	8.0	456	0.3292	0.8161	0.5
0.3585	9.0	513	0.3252	0.8161	0.5
0.3585	10.0	570	0.3175	0.8161	0.5
0.3585	11.0	627	0.3129	0.8161	0.5
0.3585	12.0	684	0.3076	0.8351	0.5029
0.3585	13.0	741	0.3024	0.8425	0.5109
0.3585	14.0	798	0.2995	0.8516	0.5163
0.3585	15.0	855	0.2953	0.8528	0.5221
0.3585	16.0	912	0.2904	0.8744	0.5426
0.3585	17.0	969	0.2875	0.8738	0.5451
0.2943	18.0	1026	0.2835	0.8833	0.5798
0.2943	19.0	1083	0.2811	0.8799	0.5710
0.2943	20.0	1140	0.2786	0.8815	0.5873
0.2943	21.0	1197	0.2761	0.8815	0.5873
0.2943	22.0	1254	0.2750	0.8838	0.5906
0.2943	23.0	1311	0.2705	0.8905	0.6194
0.2943	24.0	1368	0.2687	0.8911	0.6224
0.2943	25.0	1425	0.2674	0.8895	0.6165
0.2943	26.0	1482	0.2652	0.8911	0.6224
0.2666	27.0	1539	0.2642	0.8911	0.6224
0.2666	28.0	1596	0.2634	0.8903	0.6194
0.2666	29.0	1653	0.2612	0.8903	0.6194
0.2666	30.0	1710	0.2601	0.8921	0.6253
0.2666	31.0	1767	0.2583	0.8913	0.6328
0.2666	32.0	1824	0.2568	0.8864	0.6319
0.2666	33.0	1881	0.2563	0.8861	0.6319
0.2666	34.0	1938	0.2552	0.8869	0.6349
0.2666	35.0	1995	0.2544	0.8884	0.6378
0.2516	36.0	2052	0.2530	0.8875	0.6374
0.2516	37.0	2109	0.2523	0.8876	0.6374
0.2516	38.0	2166	0.2514	0.8889	0.6432
0.2516	39.0	2223	0.2504	0.8874	0.6453
0.2516	40.0	2280	0.2502	0.8892	0.6432
0.2516	41.0	2337	0.2495	0.8862	0.6419
0.2516	42.0	2394	0.2490	0.8867	0.6445
0.2516	43.0	2451	0.2491	0.8859	0.6365
0.2442	44.0	2508	0.2480	0.8906	0.6511
0.2442	45.0	2565	0.2476	0.8894	0.6457
0.2442	46.0	2622	0.2476	0.8888	0.6478
0.2442	47.0	2679	0.2474	0.8906	0.6511
0.2442	48.0	2736	0.2462	0.8890	0.6507
0.2442	49.0	2793	0.2461	0.8920	0.6545
0.2442	50.0	2850	0.2455	0.8894	0.6532
0.2442	51.0	2907	0.2457	0.8897	0.6507
0.2442	52.0	2964	0.2452	0.8894	0.6532
0.238	53.0	3021	0.2449	0.8903	0.6536
0.238	54.0	3078	0.2447	0.8894	0.6532
0.238	55.0	3135	0.2446	0.8894	0.6532
0.238	56.0	3192	0.2446	0.8904	0.6536
0.238	57.0	3249	0.2443	0.8894	0.6532
0.238	58.0	3306	0.2441	0.8894	0.6532
0.238	59.0	3363	0.2440	0.8911	0.6566
0.238	60.0	3420	0.2440	0.8911	0.6566
0.238	61.0	3477	0.2439	0.8903	0.6536
0.2353	62.0	3534	0.2437	0.8911	0.6566
0.2353	63.0	3591	0.2438	0.8911	0.6566
0.2353	64.0	3648	0.2437	0.8911	0.6566
0.2353	65.0	3705	0.2437	0.8911	0.6566

Framework versions

Transformers 4.41.1
Pytorch 2.3.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1

Task: Text Classification
Library: transformers
Type: mpnet
Created: May 27, 2024
Updated: Jun 5, 2024