# koelectra-small-v3-nsmc by Daekeun Kim: Open-Weight Model Source: https://savrn.com/models/koelectra-small-v3-nsmc Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve koelectra-small-v3-nsmc (14M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 23, 2026. [koelectra-small-v3-nsmc on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/koelectra-small-v3-nsmc/gpus) ## SAVRN's Notes on koelectra-small-v3-nsmc The memory line on this page reads 0.0 GB at every precision, and that is the main thing to know about koelectra-small-v3-nsmc. It is a 14M-parameter text classifier with 12 layers, about 113 MB of float32 weights on disk. The cheapest card our Index lists, one MI300X with 192 GB at $1.85 an hour, is far more than it needs, so in our facilities a model this size never gets a GPU to itself. Inputs top out at 512 tokens, which suits short text, not long documents. MIT terms permit commercial use, modification and redistribution as long as the copyright and permission notices travel with the files. Before committing, check that the nsmc data it was trained on matches the text you plan to classify. Daekeun Kim wrapped it for the SageMaker Inference Toolkit; in your own facility you write the serving layer yourself. SAVRN Research, 2026-09-18 ## Model Card By Daekeun Kim, published under mit, revision d1c0233fc013. ### Sentiment Binary Classification (fine-tuning with KoELECTRA-Small-v3 model and Naver Sentiment Movie Corpus dataset) ### Usage (Amazon SageMaker inference applicable) It uses the interface of the SageMaker Inference Toolkit as is, so it can be easily deployed to SageMaker Endpoint. #### inference_nsmc.py [Read the full model card (351 words)](https://savrn.com/models/koelectra-small-v3-nsmc/card) ## Configuration Architecture ElectraForSequenceClassification Context length (tokens) 512 Layers 12 Hidden size 256 Feed-forward size 1,024 Attention heads 4 Vocabulary size 35,000 Stored precision float32 Model type electra ## Identity and Version Repository daekeun-ml/koelectra-small-v3-nsmc Publisher Daekeun Kim Task Text classification Modality Text Library transformers Parameters 14M parameters Languages ko Revision d1c0233fc013ae8bd879d7b61f2bec69aa3af4c9 First published 2022-03-02 Last updated 2025-03-09 ## Files and Weights 10 files, 113.9 MB in total. The weights are 3 files totalling 113.1 MB in bin, safetensors. Weights3 files · 113.1 MB Configuration2 files · 1.0 KB Tokenizer3 files · 797.5 KB Documentation1 file · 5.1 KB Repository1 file · 1.2 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | model.safetensors | Weights | 56.5 MB | 012a5926dcfe | | pytorch_model.bin | Weights | 56.6 MB | 756dde30d2f3 | | training_args.bin | Weights | 1.5 KB | 6f8238eebcbe | | config.json | Configuration | 914 B | — | | special_tokens_map.json | Configuration | 112 B | — | | README.md | Documentation | 5.1 KB | — | | .gitattributes | Repository | 1.2 KB | — | | tokenizer.json | Tokenizer | 533.7 KB | — | | tokenizer_config.json | Tokenizer | 414 B | — | | vocab.txt | Tokenizer | 263.3 KB | — | ## License and Download License mit Access Open weights, no gate Download size 113.1 MB [Download from Daekeun Kim](https://huggingface.co/daekeun-ml/koelectra-small-v3-nsmc) Released by Daekeun Kim through its official repository on Hugging Face. [Read the license](https://opensource.org/license/mit). ## Built From - Trained on (disclosed) nsmc ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 113.1 MB | | 16-bit | 0.0 GB | | 8-bit | 0.0 GB | | 4-bit | 0.0 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About koelectra-small-v3-nsmc ### How much GPU memory does koelectra-small-v3-nsmc need? About 0 GB at 16-bit and 0 GB at 4-bit: the weights (14M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run koelectra-small-v3-nsmc on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### Can I use koelectra-small-v3-nsmc commercially? Yes. koelectra-small-v3-nsmc is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included. ### What is koelectra-small-v3-nsmc's context length? 512 tokens, from the maximum position embeddings in its published configuration. ## Similar Models Model · Text classification ### [koelectra-small-v3-discriminator](https://savrn.com/models/koelectra-small-v3-discriminator) [Kim MinKyung](https://savrn.com/model-publishers/minkyungeee) Open weights 14M parameters transformers [View model](https://savrn.com/models/koelectra-small-v3-discriminator) Model · Text classification ### [koelectra-base-v3-discriminator-news-classifier](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-14) [Sonjunsu](https://savrn.com/model-publishers/hwirang) This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019). Open weights 14M parameters 512 tokens transformers [View model](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-14) Model · Text classification ### [koelectra-base-v3-discriminator-news-classifier](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier) [Bae Suk Geun](https://savrn.com/model-publishers/sgbae) Open weights 14M parameters transformers [View model](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier) Model · Text classification ### [koelectra-base-v3-discriminator-news-classifier](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-2) [Lim Jae Hyoung](https://savrn.com/model-publishers/ljh6625) Open weights 14M parameters transformers [View model](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-2) Model · Text classification ### [koelectra-base-v3-discriminator-news-classifier](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-3) [Sea](https://savrn.com/model-publishers/yyeeoon031) Open weights 14M parameters transformers [View model](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-3) Model · Text classification ### [koelectra-base-v3-discriminator-news-classifier](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-4) [Sanghwan Ryu](https://savrn.com/model-publishers/shryuai0075) Open weights 14M parameters transformers [View model](https://savrn.com/models/koelectra-base-v3-discriminator-news-classifier-4) ## Daekeun Kim [All models and datasets](https://savrn.com/model-publishers/daekeun-ml) ## Versions - [d1c0233fc013](https://savrn.com/models/koelectra-small-v3-nsmc/versions/d1c0233fc013) · current 2026-09-27 ## Explore More - [All text classification models](https://savrn.com/models/tasks/text-classification) - [All models under mit](https://savrn.com/models/licenses/mit) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-19. - [Hugging Face record](https://huggingface.co/daekeun-ml/koelectra-small-v3-nsmc) - [How the hub is built](https://savrn.com/model-hub/methodology)