starlinghq 

Managed .NET tokenizer for Hugging Face tokenizer.json files (format 1.0): Unigram, byte-level BPE, SentencePiece-style BPE with byte fallback, and WordPiece. Exact against huggingface/tokenizers on published... More information

Explicit, verified acquisition of model assets: downloads the files of an immutable asset lock (Hub repository, full commit SHA, per-file size and SHA-256) into a local cache, verifies them, and resolves them... More information

Text generation on ONNX Runtime GenAI 0.17.1 behind a Starling adapter: model loading with explicit provider selection (no silent fallback, TF32 off by default), runtime config overlays, reference-counted... More information

Qwen/Qwen3.5-0.8B (revision 2fc06364) chat generation in-process on ONNX Runtime GenAI 0.17.1: the pinned chat template rendered in C# for text and tool-calling conversations (system, user, assistant and tool... More information

ONNX Runtime session ownership, execution-provider validation and diagnostics for Starlings models. References only the managed ONNX Runtime binding (exactly 1.30.0): the application adds one native package,... More information

google/embeddinggemma-2 (revision 914f7f89) text and image embeddings on ONNX Runtime, from text through the owned tokenizer (Starlings.Tokenizers, profile gemma-bpe-embeddinggemma2: no Microsoft.ML.Tokenizers,... More information

intfloat/multilingual-e5-small (revision 614241f6) sentence embeddings on ONNX Runtime, from text through the owned tokenizer (Starlings.Tokenizers: no Microsoft.ML.Tokenizers, no native tokenizer) or from... More information

Microsoft.Extensions.AI IEmbeddingGenerator<string, Embedding<float>> over the multilingual-e5-small text embedder (Starlings.Models.E5): explicit e5 input-type prefixes, batching, usage, node-boundary... More information

Microsoft.Extensions.AI IChatClient over the Qwen/Qwen3.5-0.8B chat model (Starlings.Models.Qwen35): system, user and assistant text messages; tool calling (AIFunction tools, FunctionCallContent and... More information

Microsoft.Extensions.AI IEmbeddingGenerator<string, Embedding<float>> over the EmbeddingGemma 2 text embedder (Starlings.Models.EmbeddingGemma2): explicit named prompts, Dimensions mapped to the Matryoshka... More information

fastino/gliner2-base-v1 (revision f9634218, GLiNER2 2.0.0) English entity extraction with labels and a threshold on ONNX Runtime: GLiNER2's Python text semantics (normalization, word splitter, lower-casing) and... More information

ML.NET estimator and transformer for the multilingual-e5-small text embedder (Starlings.Models.E5): a text column to a fixed-size Vector<Single, 384> embedding column and an optional token-count column, with... More information

ML.NET estimator and transformer for the EmbeddingGemma 2 text embedder (Starlings.Models.EmbeddingGemma2): a text column to a fixed-size Vector<Single, 768/512/256/128> embedding column and an optional... More information

Decodes image files and streams (PNG, JPEG, GIF, BMP and WebP experimental; GIF with transparency rejected) with SkiaSharp into Starlings.Models.EmbeddingGemma2 images, reproducing PIL's convert("RGB") as the... More information

The FFmpeg 9 counterpart of Starlings.Models.EmbeddingGemma2.Video.FFmpeg: decodes video files and streams into Starlings.Models.EmbeddingGemma2 videos the way upstream does (torchcodec 0.17 on FFmpeg):... More information

Decodes video files and streams into Starlings.Models.EmbeddingGemma2 videos the way upstream does (torchcodec 0.17 on FFmpeg): torchcodec's packet-scan metadata, upstream's 1 frame per second sampling (at most... More information

Qwen/Qwen3-0.6B (revision c1899de2, thinking disabled) and Qwen/Qwen2.5-0.5B-Instruct (revision 7ae55760) text chat generation in-process on ONNX Runtime GenAI 0.17.1: the pinned chat templates rendered in C#... More information

Kev decision models (jaredpalmer/kev-0.8b and jaredpalmer/kev-4b, Kev 1.0) on ONNX Runtime: one state and typed questions (noul, choice, score) in, a calibrated probability per option out, in one prefill pass... More information