Ornith-1.5-35B-ROCmFP4-STRIX_LEAN
Copyright 2026 pugant

This is a derivative work. Original model and its license are preserved.

DERIVATION CHAIN:
  Qwen3.5-VL-MoE (base architecture)
    └── ornith-ai/Ornith-1.5-35B (A3B) (MIT License, Copyright DeepReinforce Team)
            └── this GGUF quantization by pugant

QUANTIZATION:
  Preset: Q4_0_ROCMFP4_STRIX_LEAN (type 106, ~4.29 BPW)
  Tool: charlie12345/ROCmFPX fork of llama.cpp (MIT)
        https://github.com/charlie12345/ROCmFPX
  Source BF16 GGUF: ornith-ai/Ornith-1.5-35B-A3B-GGUF (Ornith-1.5-35B-BF16.gguf, MIT)
  imatrix: bartowski (precomputed on 573 chunks, calibration-v6,
           included as imatrix-Ornith-1.5-35B-bartowski.gguf with attribution)
           https://huggingface.co/bartowski/Ornith-1.5-35B-A3B-GGUF

RUNTIME:
  Requires kyuz0/amd-strix-halo-toolboxes + charlie12345/ROCmFPX
  https://github.com/kyuz0/amd-strix-halo-toolboxes