RecursiveAI.OneNlp.Models.Hindi
1.0.0
Prefix Reserved
dotnet add package RecursiveAI.OneNlp.Models.Hindi --version 1.0.0
NuGet\Install-Package RecursiveAI.OneNlp.Models.Hindi -Version 1.0.0
<PackageReference Include="RecursiveAI.OneNlp.Models.Hindi" Version="1.0.0" />
<PackageVersion Include="RecursiveAI.OneNlp.Models.Hindi" Version="1.0.0" />
<PackageReference Include="RecursiveAI.OneNlp.Models.Hindi" />
paket add RecursiveAI.OneNlp.Models.Hindi --version 1.0.0
#r "nuget: RecursiveAI.OneNlp.Models.Hindi, 1.0.0"
#:package RecursiveAI.OneNlp.Models.Hindi@1.0.0
#addin nuget:?package=RecursiveAI.OneNlp.Models.Hindi&version=1.0.0
#tool nuget:?package=RecursiveAI.OneNlp.Models.Hindi&version=1.0.0
RecursiveAI.OneNLP.Models.Hindi
spaCy-class NLP for .NET. The Hindi fast-tier model for RecursiveAI.OneNlp: add this package and the Nlp verbs analyze Hindi text with part-of-speech tags, morphological features, a dependency parse and named entities (PER, ORG, LOC). No downloads, no native code, NativeAOT and trim clean. Unaffiliated with other projects named OneNLP.
Weaker than the other languages' models. There is no Hindi treebank whose licence allows training a model that is shipped (Hindi-HDTB is non-commercial). This model was trained on Hindi web text (FineWeb-2) that the multilingual accurate-tier model labelled by cross-lingual transfer from other languages (Marathi, Sindhi and five large languages), keeping the sentences where its sources agree. Its labels are therefore noisier than a treebank's, and its scores on the UD Hindi-PUD test set (gold tokens: UPOS 83.2, LAS 52.8) are lower than the English or Spanish models' on their treebanks. It has no lemmatizer:
Token.Lemmais the token's own text, which is wrong for inflected words, and there are no XPOS tags (Token.Tagis empty), because no transferable source for them exists. The entity recognizer was never scored against Hindi entity gold, because none that may be used exists; seePROVENANCE.mdfor the measurements that were possible.
using RecursiveAI.OneNlp;
Doc doc = Nlp.Analyze("प्रधानमंत्री नरेंद्र मोदी ने दिल्ली में कल एक बैठक की। उन्होंने कहा कि सरकार तैयार है।");
foreach (Token t in doc)
{
Console.WriteLine($"{t} {t.Pos} {t.Dep}");
}
foreach (Span e in Nlp.Entities("नरेंद्र मोदी दिल्ली में हैं।"))
{
Console.WriteLine($"{e.Label} {e}");
}
Words are separated by spaces and the danda । is a token of its own; Hindi has no multiword tokens. The package adds a small source file to the project that references it; that file registers the model when the program starts. Programs that cannot compile it call HindiModel.Register() once.
The weights are derived from Hindi web text (FineWeb-2, ODC-By 1.0, Common Crawl terms) and from the accurate-tier model, which was trained on Universal Dependencies treebanks (CC BY-SA and CC BY) and named-entity corpora; every source is credited in THIRD-PARTY-NOTICES.md, and PROVENANCE.md lists every dataset, recipe and seed. The weights are licensed MIT OR Apache-2.0.
- Measured results: https://github.com/recursive-ai-oss/onenlp-csharp/blob/main/docs/results-table.md
- Source and documentation: https://github.com/recursive-ai-oss/onenlp-csharp
- Home: https://recursiveai.org/onenlp/
Copyright (c) Recursive AI, LLC.
| Product | Versions Compatible and additional computed target framework versions. |
|---|---|
| .NET | net10.0 is compatible. net10.0-android was computed. net10.0-browser was computed. net10.0-ios was computed. net10.0-maccatalyst was computed. net10.0-macos was computed. net10.0-tvos was computed. net10.0-windows was computed. |
-
net10.0
- RecursiveAI.OneNlp.Neural (>= 1.0.0)
NuGet packages
This package is not used by any NuGet packages.
GitHub repositories
This package is not used by any popular GitHub repositories.
| Version | Downloads | Last Updated |
|---|---|---|
| 1.0.0 | 116 | 11/8/2026 |
See CHANGELOG.md in the repository.