TSE_XLNET_5E

Etelis/TSE_XLNET_5E

TSE_XLNET_5E is a text classification model from Etelis. Use it when you need a label for a piece of text. 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

39% of all-time downloads

All-time downloads

Public

Repo size

939 MB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin469 MB · 99%

At a glance

Task: Text Classification
Library: transformers
License: mit
Model type: xlnet
Access: Public
Created: Nov 9, 2022
Updated: Nov 9, 2022
SHA: a5d16f80

Try a prompt

Model card

From the Hugging Face model README

TSE_XLNET_5E

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

Loss: 0.4463
Accuracy: 0.9333

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: 3e-05
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6717	0.06	50	0.4377	0.8533
0.3989	0.12	100	0.4525	0.84
0.3433	0.17	150	0.3348	0.9133
0.2646	0.23	200	0.3722	0.9
0.3052	0.29	250	0.3306	0.8933
0.2583	0.35	300	0.3129	0.92
0.2712	0.4	350	0.3147	0.9
0.2708	0.46	400	0.2680	0.9
0.2443	0.52	450	0.2261	0.9133
0.2463	0.58	500	0.2583	0.9067
0.2525	0.63	550	0.2719	0.92
0.2522	0.69	600	0.3905	0.8933
0.2078	0.75	650	0.2674	0.9133
0.264	0.81	700	0.2774	0.9133
0.211	0.87	750	0.2652	0.9333
0.286	0.92	800	0.1777	0.94
0.2341	0.98	850	0.2570	0.9133
0.1797	1.04	900	0.3162	0.92
0.1831	1.1	950	0.3205	0.92
0.2006	1.15	1000	0.3173	0.9133
0.1555	1.21	1050	0.3388	0.9267
0.1712	1.27	1100	0.3968	0.92
0.1488	1.33	1150	0.4167	0.9133
0.1893	1.38	1200	0.3269	0.9267
0.1543	1.44	1250	0.3797	0.9133
0.1825	1.5	1300	0.2203	0.94
0.1841	1.56	1350	0.2744	0.9133
0.1523	1.61	1400	0.3561	0.9067
0.1914	1.67	1450	0.2859	0.9067
0.1742	1.73	1500	0.2461	0.9267
0.145	1.79	1550	0.4266	0.9133
0.208	1.85	1600	0.3470	0.9067
0.147	1.9	1650	0.4521	0.9133
0.1867	1.96	1700	0.3648	0.9067
0.182	2.02	1750	0.2659	0.9333
0.1079	2.08	1800	0.3393	0.92
0.1338	2.13	1850	0.3483	0.9267
0.1181	2.19	1900	0.4384	0.92
0.1418	2.25	1950	0.3468	0.9267
0.0953	2.31	2000	0.4008	0.9267
0.1313	2.36	2050	0.3301	0.9333
0.0499	2.42	2100	0.4018	0.92
0.1197	2.48	2150	0.3394	0.9267
0.1237	2.54	2200	0.3399	0.92
0.0766	2.6	2250	0.3947	0.9267
0.1142	2.65	2300	0.4055	0.9133
0.1362	2.71	2350	0.2599	0.9333
0.1332	2.77	2400	0.3293	0.9133
0.1241	2.83	2450	0.3717	0.92
0.0696	2.88	2500	0.4440	0.92
0.1012	2.94	2550	0.4026	0.92
0.1028	3.0	2600	0.4202	0.9133
0.0551	3.06	2650	0.4649	0.9133
0.0796	3.11	2700	0.4053	0.92
0.0786	3.17	2750	0.4862	0.9067
0.0843	3.23	2800	0.4007	0.9267
0.0502	3.29	2850	0.4510	0.92
0.0726	3.34	2900	0.4171	0.9267
0.0933	3.4	2950	0.3485	0.9333
0.0624	3.46	3000	0.4442	0.9133
0.0475	3.52	3050	0.4449	0.92
0.0498	3.58	3100	0.4147	0.9267
0.1101	3.63	3150	0.3484	0.9333
0.0785	3.69	3200	0.3630	0.9267
0.075	3.75	3250	0.4267	0.92
0.0709	3.81	3300	0.3638	0.9267
0.0754	3.86	3350	0.3890	0.9333
0.1038	3.92	3400	0.3910	0.9267
0.0274	3.98	3450	0.4246	0.9267
0.0723	4.04	3500	0.3847	0.9267
0.015	4.09	3550	0.4134	0.9333
0.0329	4.15	3600	0.4136	0.9333
0.0619	4.21	3650	0.4048	0.9333
0.0505	4.27	3700	0.4228	0.9267
0.0523	4.33	3750	0.4139	0.9267
0.0365	4.38	3800	0.4067	0.9267
0.0434	4.44	3850	0.4132	0.9333
0.0262	4.5	3900	0.4245	0.9333
0.0534	4.56	3950	0.4217	0.9333
0.0186	4.61	4000	0.4282	0.9333
0.0548	4.67	4050	0.4255	0.9333
0.0146	4.73	4100	0.4368	0.9333
0.0442	4.79	4150	0.4470	0.9333
0.0431	4.84	4200	0.4469	0.9333
0.0297	4.9	4250	0.4470	0.9333
0.0601	4.96	4300	0.4463	0.9333

Framework versions

Transformers 4.24.0
Pytorch 1.13.0
Datasets 2.3.2
Tokenizers 0.13.1

Task: Text Classification
Library: transformers
Type: xlnet
License: mit
Created: Nov 9, 2022
Updated: Nov 9, 2022