topic-classifier-finetuned

epreep/topic-classifier-finetuned

topic-classifier-finetuned is a text classification model from epreep. Use it when you need a label for a piece of text. 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

11% of all-time downloads

All-time downloads

Public

Parameters

111M

443 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors443 MB · 100%

At a glance

Task: Text Classification
Library: transformers
License: apache-2.0
Model type: roberta
Access: Public
Created: Sep 22, 2025
Updated: Sep 22, 2025
SHA: 9f81a35b

Try a prompt

Base models

pongjin/roberta_with_kornli

Model card

From the Hugging Face model README

topic-classifier-finetuned

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

Loss: 0.2698
Accuracy: 0.9271
F1 Macro: 0.9270

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: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 64
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro
2.6509	0.0730	100	1.6595	0.7331	0.7135
1.1192	0.1459	200	0.6936	0.8678	0.8659
0.6712	0.2189	300	0.5144	0.8881	0.8881
0.5925	0.2919	400	0.4652	0.8915	0.8921
0.515	0.3648	500	0.4157	0.9000	0.8999
0.4675	0.4378	600	0.4020	0.8990	0.8990
0.4408	0.5108	700	0.3746	0.9039	0.9038
0.4237	0.5837	800	0.3597	0.9034	0.9041
0.4147	0.6567	900	0.3420	0.9057	0.9054
0.3874	0.7297	1000	0.3167	0.9121	0.9118
0.3614	0.8026	1100	0.3415	0.9081	0.9073
0.3651	0.8756	1200	0.3207	0.9097	0.9098
0.326	0.9486	1300	0.3178	0.9147	0.9142
0.3455	1.0212	1400	0.3235	0.9127	0.9120
0.2684	1.0941	1500	0.3038	0.9151	0.9150
0.2593	1.1671	1600	0.3101	0.9127	0.9121
0.2639	1.2401	1700	0.2992	0.9144	0.9147
0.2595	1.3130	1800	0.3078	0.9146	0.9144
0.2681	1.3860	1900	0.2959	0.9156	0.9157
0.2578	1.4590	2000	0.2909	0.9187	0.9183
0.2555	1.5319	2100	0.3025	0.9155	0.9149
0.2581	1.6049	2200	0.2815	0.9203	0.9201
0.2478	1.6779	2300	0.2833	0.9219	0.9216
0.2428	1.7508	2400	0.2831	0.9203	0.9202
0.2638	1.8238	2500	0.2710	0.9249	0.9248
0.2462	1.8968	2600	0.2799	0.9209	0.9208
0.2526	1.9697	2700	0.2826	0.9187	0.9189
0.2147	2.0423	2800	0.2718	0.9242	0.9241
0.1757	2.1153	2900	0.2817	0.9248	0.9248
0.1727	2.1883	3000	0.2821	0.9237	0.9235
0.1836	2.2612	3100	0.2875	0.9209	0.9211
0.1657	2.3342	3200	0.2767	0.9249	0.9248
0.1708	2.4072	3300	0.2757	0.9237	0.9237
0.1693	2.4801	3400	0.2752	0.9233	0.9233
0.1836	2.5531	3500	0.2793	0.9225	0.9224
0.1651	2.6260	3600	0.2790	0.9237	0.9236
0.1675	2.6990	3700	0.2741	0.9247	0.9247
0.1661	2.7720	3800	0.2717	0.9264	0.9263
0.1681	2.8449	3900	0.2718	0.9249	0.9249
0.1672	2.9179	4000	0.2695	0.9273	0.9271
0.1693	2.9909	4100	0.2698	0.9271	0.9270

Framework versions

Transformers 4.56.2
Pytorch 2.8.0+cu128
Datasets 4.1.1
Tokenizers 0.22.1

Task: Text Classification
Library: transformers
Type: roberta
License: apache-2.0
Created: Sep 22, 2025
Updated: Sep 22, 2025