Module api 

GGUF high-level API (load, tokenizer, model config)

Structs

GgufLoadResult

Result of loading a GGUF file with full tokenizer data and model config

GgufModelConfig

Transformer model configuration extracted from GGUF metadata CRITICAL for APR inference - must match realizar::apr::AprMetadata

GgufRawLoadResult

Result of loading GGUF with raw quantized tensors

GgufRawTensor

Raw tensor data with quantization preserved

GgufTokenizer

Tokenizer data extracted from GGUF file

Functions

load_gguf_raw

Load GGUF with raw quantized tensors (preserves Q4K for GPU inference)

load_gguf_tensors

Load GGUF file and extract all tensors as F32

load_gguf_with_tokenizer

Load GGUF file and extract tensors, tokenizer, AND model config