Group: Qxotic

Sort by:Popular▼

✓Popular✓Newest

Core model APIs, execution, memory management, and telemetry for Jinfer.

Memory allocation, views, access, and transfer APIs for Jota.

Foundational Jota data types: devices, shapes, strides, and layouts.

Customizable tiktoken-compatible and SentencePiece BPE tokenization for language models.

Java library for reading and writing GGUF metadata and tensor descriptors for llama.cpp model files.

Chat messages, prompt templates, and structured response parsing for Jinfer.

Compute kernels and GGUF tensor loading for Jinfer model providers.

Loads Tok'n'Roll tokenizers from GGUF model files.

Strict RFC 8259-compliant JSON parser and writer for Java.

Media decoding for Jinfer image, audio, and video inputs, and espeak-ng grapheme-to-phoneme, on the JVM and in native images.

Core API for quantized matrix multiplication with JAM.

Persistent, memory-bounded prompt caching for Jinfer models.

Resolves and downloads GGUF model files from Hugging Face and ModelScope.

Spring AI provider for local AI inference with Jinfer. Supports chat, streaming, embeddings, reranking, tools, multimodal input, and speech.

Local chat inference for inclusionAI Ling 3 models (Bailing V3 MoE architecture) with Jinfer.

Local chat, vision, and audio inference for Google Gemma 4 models with Jinfer.

Local chat inference for OpenAI gpt-oss models with Jinfer.

Local text-to-speech inference for Owen Song's Inflect 2 models with Jinfer.

Standalone Jinja chat-template rendering for Java.

Native Kokoro text-to-speech inference with Jinfer.