---

license: other

license_name: llama4

base_model: meta-llama/Llama-4-Scout-17B-16E-Instruct

base_model_relation: quantized

tags: [gguf, llama.cpp, rocm, gfx1151, strix-halo, amd, ryzen-ai-max-395, rocmfp4, rocmfpx, llama4, llama-4, scout, moe, vision, multimodal]

language: [en]

pipeline_tag: image-text-to-text

---

> ### ⚠️ STOCK llama.cpp WILL NOT LOAD THIS MODEL

>

> Q4_0_ROCMFP4_COHERENT (ftype 102) exists only in

> charlie12345/ROCmFPX. Stock llama.cpp reports

> invalid ggml type. Ignore the auto-generated "Use this model" commands above.

>

> ⚠️ -fa off is required — flash attention breaks the vision path on gfx1151.

>

> 56.98 GiB · 17.72 tok/s on a Ryzen AI MAX+ 395 (Strix Halo, gfx1151).

Llama-4-Scout-17B-16E-Instruct — ROCmFP4 (tier 102 COHERENT) GGUF

The first ROCmFP4 build of Llama 4 Scout. 109B total / 17B active MoE with vision, quantized

for AMD gfx1151 (Ryzen AI MAX+ 395 / Strix Halo).

| | |

|---|---|

| File | Llama-4-Scout-17B-16E-Instruct-Q4_0_ROCMFP4_COHERENT.gguf |

| Size | 56.98 GiB (61,182,279,392 B) |

| BPW | 4.54 |

| ftype | Q4_0_ROCMFP4_COHERENT (102) |

| Tensors | 628 · 480 at q4_0_rocmfp4 |

| Architecture | 48 layers · hidden 5120 · 16 experts, 1 active/token · vision patch 14 |

Quantized from unsloth's BF16 GGUF (200.8 GiB, 5 shards, byte-verified against the Hub) — a

lossless source, not a requantization and not our own re-conversion.

Recipe

--output-tensor-type q6_K --token-embedding-type q6_K   ...   Q4_0_ROCMFP4_COHERENT 16

⭐ --output-tensor-type q6_K is what actually protects the LM head — the COHERENT tier

alone still leaves output.weight at 4-bit. Audited in the finished file: output.weight q6_K,

token_embd.weight q6_K.

---

Measured

Ryzen AI MAX+ 395, gfx1151, ROCm. -ngl 999 -c 4096 -fa off -fit off --mmproj.

Median of 3, warm-up discarded, otherwise-idle box.

| | |

|---|---|

| Decode | 17.72 tok/s (17.69 / 17.72 / 17.74 — tight) |

| Load | 48 s cold, 22 s warm |

Correctness (max_tokens: 1024): 17×23 ⇒ ✅ 391 · capital of Japan ⇒ ✅ Tokyo ·

days in 2024 ⇒ ✅ 366

Vision (-fa off + bundled BF16 mmproj):

| image | result |

|---|---|

| 256×256 solid red | ✅ Red. |

| 256×256 shapes | ✅ red square · green circle · blue triangle · yellow border · black background |

| 512×512 | ⛔ failed to encode image slice — server dies |

⚠️ Use images at or near 256×256. 512×512 fails to encode on this build. The projector is

fine — 256×256 reads colour and shape and spatial layout correctly — but larger images hit an

encode-slice limit. Resize before sending.

No draft head. This model ships no MTP/EAGLE/nextn tensors

(common_speculative_init: no implementations specified), so there is no speculative-decoding

speedup to enable and no acceptance-length figure to quote.

---

What was NOT measured

• No perplexity run, and no quality A/B against the BF16 source.
• No long-context testing. · No tool-calling evaluation.
• Vision was smoke-tested on synthetic images only — no vision benchmark.

Base model licence (Llama 4 Community License) inherited; all credit for the weights goes to Meta.