# piiranha-v1-detect-personal-information by The Internet Source: https://savrn.com/models/piiranha-v1-detect-personal-information Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve piiranha-v1-detect-personal-information (278M 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 | 0.6 GB | 0.7 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 | 0.3 GB | 0.3 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 | 0.1 GB | 0.2 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 22, 2026. [piiranha-v1-detect-personal-information on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/piiranha-v1-detect-personal-information/gpus) ## SAVRN's Notes on piiranha-v1-detect-personal-information Piiranha detects 17 types of personal information across six languages, and its authors report catching more than 98 percent of PII tokens in their test. It is a 278M-parameter DeBERTa model. The license is CC-BY-NC-ND-4.0: no commercial use and no derivatives. For commercial redaction, OpenAI's privacy filter and OpenMed's PII models in our catalog carry Apache 2.0 licenses. SAVRN Research, 2026-09-18 ## Model Card william (at) integrinet [dot] org Piiranha (cc-by-nc-nd-4.0 license) is trained to detect 17 types of Personally Identifiable Information (PII) across six languages. It successfully catches 98.27% of PII tokens, with an overall classification accuracy of 99.44%. Piiranha is especially accurate at detecting passwords, emails (100%), phone numbers, and usernames. Performance on PII vs. Non PII classification task: Piiranha was trained on H100 GPUs generously sponsored by the Akash Network Piiranha is a fine-tuned version of microsoft/mdeberta-v3-base. The context length is 256 Deberta tokens. If your text is longer than that, just split it up. Supported PII types: Account Number, Building… Excerpt from the card by The Internet Integrity Initiative, licensed cc-by-nc-nd-4.0. ## Configuration Architecture DebertaV2ForTokenClassification Context length (tokens) 512 Layers 12 Hidden size 768 Feed-forward size 3,072 Attention heads 12 Vocabulary size 251,000 Stored precision float32 Model type deberta-v2 ## Identity and Version Repository iiiorg/piiranha-v1-detect-personal-information Publisher The Internet Integrity Initiative Task Token classification Modality Text Library transformers Parameters 278M parameters Languages en, it, fr, de, nl, es Revision 255acde67a2f34cf452eb42e365b24d2957352fc First published 2024-09-12 Last updated 2025-11-01 ## Files and Weights 10 files, 1.1 GB in total. The weights are 2 files totalling 1.1 GB in bin, safetensors. Weights2 files · 1.1 GB Configuration3 files · 2.1 KB Tokenizer2 files · 16.4 MB Documentation1 file · 9.0 KB Other1 file · 4.3 MB Repository1 file · 1.6 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | model.safetensors | Weights | 1.1 GB | 3d503b6804dd | | training_args.bin | Weights | 5.1 KB | ffd4943d9df1 | | added_tokens.json | Configuration | 23 B | — | | config.json | Configuration | 1.7 KB | — | | special_tokens_map.json | Configuration | 286 B | — | | README.md | Documentation | 9.0 KB | — | | spm.model | Other | 4.3 MB | 13c8d666d62a | | .gitattributes | Repository | 1.6 KB | — | | tokenizer.json | Tokenizer | 16.4 MB | 9ddbe35b768b | | tokenizer_config.json | Tokenizer | 19.7 KB | — | ## License and Download License cc-by-nc-nd-4.0 Access Open weights, no gate Download size 1.1 GB [Download from The Internet Integrity Initiative](https://huggingface.co/iiiorg/piiranha-v1-detect-personal-information) Released by The Internet Integrity Initiative through its official repository on Hugging Face. ## Built From - Derived from [microsoft/mdeberta-v3-base](https://savrn.com/models/mdeberta-v3-base) - Trained on (disclosed) ai4privacy/pii-masking-400k ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 1.1 GB | | 16-bit | 0.6 GB | | 8-bit | 0.3 GB | | 4-bit | 0.1 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About piiranha-v1-detect-personal-information ### How much GPU memory does piiranha-v1-detect-personal-information need? About 0.7 GB at 16-bit and 0.2 GB at 4-bit: the weights (278M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run piiranha-v1-detect-personal-information 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. ### What license is piiranha-v1-detect-personal-information released under? cc-by-nc-nd-4.0, as its publisher declares it. Read the license text before commercial use. ### What is piiranha-v1-detect-personal-information's context length? 512 tokens, from the maximum position embeddings in its published configuration. ## Similar Models Model · Token classification ### [OpenMed-NER-OncologyDetect-BigMed-278M](https://savrn.com/models/openmed-ner-oncologydetect-bigmed-278m) [OpenMed](https://savrn.com/model-publishers/openmed) Specialized model for Cancer Genetics - Cancer-related genetic entities This model is a state-of-the-art fine-tuned transformer engineered to deliver enterprise-grade accuracy for cancer genetics - cancer-related genetic entities. This specialized model excels at identifying and extracting biomedical entities from clinical texts, research papers, and healthcare documents, enabling applications such as drug interaction detection, medication extraction from patient records, adverse event monitoring, literature mining for drug discovery, and biomedical knowledge graph construction with production-ready reliability for clinical and research applications. This model can identify and classify the… Open weights apache-2.0 277M parameters 514 tokens transformers [View model](https://savrn.com/models/openmed-ner-oncologydetect-bigmed-278m) Model · Token classification ### [actividad10-ner-xlmr](https://savrn.com/models/actividad10-ner-xlmr) [Yorquin Yobany Rojas Silva](https://savrn.com/model-publishers/yobanyr) Open weights 277M parameters [View model](https://savrn.com/models/actividad10-ner-xlmr) Model · Token classification ### [halluvision-cyner-xlmr-base](https://savrn.com/models/halluvision-cyner-xlmr-base) [Seungho Song](https://savrn.com/model-publishers/ricky-dev) Token-classification checkpoint used to produce the reference entities in It is CyNER (Alam et al., 2022) fine-tuned on 50 MITRE ATT&CK software articles annotated with five STIX-aligned labels. Every entity-based number in the paper (answer-matching precision/recall, frequency-based entity similarity, the Threatpost and The Hacker News ground truth, the Llama-3 summary NER) was produced by this exact checkpoint, so it is released to make those numbers regenerable. BIO tagging over five entity types: Label definitions follow Table 3 of the paper (Malware, System, Indicator, Vulnerability, Organization). Fine-tuned with the CyNER training loop (cyner.TransformersNER) from xlm-roberta-base.… Open weights mit 277M parameters 514 tokens transformers [View model](https://savrn.com/models/halluvision-cyner-xlmr-base) Model · Token classification ### [xlm-roberta-base-ner-hrl](https://savrn.com/models/xlm-roberta-base-ner-hrl) [David Adelani](https://savrn.com/model-publishers/davlan) Hugging Face's logo - multilingual xlm-roberta-base-ner-hrl is a Named Entity Recognition model for 10 high resourced languages (Arabic, German, English, Spanish, French, Italian, Latvian, Dutch, Portuguese and Chinese) based on a fine-tuned XLM-RoBERTa base model. It has been trained to recognize three types of entities: location (LOC), organizations (ORG), and person (PER). Specifically, this model is a xlm-roberta-base model that was fine-tuned on an aggregation of 10 high-resourced languages You can use this model with Transformers pipeline for NER. This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well… Open weights afl-3.0 277M parameters 514 tokens transformers [View model](https://savrn.com/models/xlm-roberta-base-ner-hrl) Model · Token classification ### [OpenMed-NER-PharmaDetect-BigMed-278M](https://savrn.com/models/openmed-ner-pharmadetect-bigmed-278m) [OpenMed](https://savrn.com/model-publishers/openmed) Specialized model for Chemical Entity Recognition - Chemical entities from the BC5CDR dataset This model is a state-of-the-art fine-tuned transformer engineered to deliver enterprise-grade accuracy for chemical entity recognition - chemical entities from the bc5cdr dataset. This specialized model excels at identifying and extracting biomedical entities from clinical texts, research papers, and healthcare documents, enabling applications such as drug interaction detection, medication extraction from patient records, adverse event monitoring, literature mining for drug discovery, and biomedical knowledge graph construction with production-ready reliability for clinical and research… Open weights apache-2.0 277M parameters 514 tokens transformers [View model](https://savrn.com/models/openmed-ner-pharmadetect-bigmed-278m) Model · Token classification ### [intenter](https://savrn.com/models/intenter) [Elda](https://savrn.com/model-publishers/elda-ai) Korean-first, multilingual (ko · ja · en). 281M parameters · p50 27.8 ms. Elda-Intenter reads an utterance and returns, in one forward pass, four aligned channels: what the speaker is doing (speech act), what they are talking about (entities), and the attribute/predicate structure that binds them. It is not a general-purpose NER model. It is the perception layer of a live conversational system, built for the constraint that shapes everything else in that setting: latency budgets measured in tens of milliseconds. All figures below are on sets excluded from training by construction. KLUE-NER (dev, Korean · 600 sentences / 1,687 gold entities), entity-level F1: Korean particles attach to the… Access requested at publisher other 281M parameters transformers [View model](https://savrn.com/models/intenter) ## The Internet Integrity Initiative [All models and datasets](https://savrn.com/model-publishers/iiiorg) ## Versions - [255acde67a2f](https://savrn.com/models/piiranha-v1-detect-personal-information/versions/255acde67a2f) · current 2026-09-18 ## Explore More - [All token classification models](https://savrn.com/models/tasks/token-classification) - [All models under cc-by-nc-nd-4.0](https://savrn.com/models/licenses/cc-by-nc-nd-4-0) - [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 2026-10-20. - [Hugging Face record](https://huggingface.co/iiiorg/piiranha-v1-detect-personal-information) - [How the hub is built](https://savrn.com/model-hub/methodology)