w11wo / token-classification updated 2 years ago

indonesian-roberta-base-posp-tagger

Name: indonesian-roberta-base-posp-tagger
Author: w11wo

This model is a fine-tuned version of flax-community/indonesian-roberta-base on the indonlu dataset. It achieves the following results on the evaluation set: - Loss: 0.1395 - Precision: 0.9625 - Recall: 0.9625 - F1: 0.9625 - Accuracy: 0.9625

Params

120 M

Context

512

Downloads 30d

3.1 M

Likes

Commercial use: allowed mit Not gated SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

3.2 M3.1 M

Jul 28Jul 31Aug 3Aug 6

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	f32	0.5 GB	1.1 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Architecture: RobertaForTokenClassification
Parameters: 120 M
Tensor type: F32
Context length: 512
Vocabulary: 50,265
Layers / heads: 12 / 12
Licence: mit
First seen on the Hub: 2022-03-02
Training datasets: indonlu
indonlu (reported): 0.96251002405774
Added to our catalog: 2026-07-28

Compare with

stanford-deidentifier-base 1.9 M bert-base-NER 1.3 M punctuate-all 708 K ner-english-fast 543 K llmlingua-2-bert-base-multilingual-cased-meetingbank 493 K bert-portuguese-ner 452 K

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB