loom-fun/ggml-org-Flint-loom

Flint (Q4_K_M) for AMD Loom

A Loom package (.loom programs and a model.json card) for the unmodified GGUF checkpoint ggml-org/Flint-GGUF (Flint-Q4_K_M.gguf), generated on an AMD Ryzen AI Max+ 395 "Strix Halo" with the loom-skills workflow.

Quick start

pip install "hrx-loom[hf]"
loom-fetch https://huggingface.co/loom-fun/ggml-org-Flint-loom/resolve/main/model.json --materialize=copy --output=flint
loom-serve --model=flint/model.json --device=amdgpu --memory_bytes=67000000000

The server listens on 127.0.0.1:8080 and answers OpenAI-style /v1/chat/completions requests.

Performance vs llama.cpp

metric	llama.cpp	Loom	margin	roofline	Loom % of roofline	Loom on DevLab (% of roofline)
decode tok/s, short prompt	61.0	98.1	+61 %	144.5	68 %	96.9 (67 %)
decode tok/s, 2k context	54.1	84.9	+57 %	146.4	58 %	65.9 (45 %)
decode tok/s, 32k context	46.8	45.1	-4 %	98.9	46 %	44.1 (45 %)
prefill tok/s, 1k document	1046.2	912.1	-13 %	12248.4	7 %	701.6 (6 %)
prefill tok/s, 2k document	1060.0	1287.1	+21 %	11945.5	11 %	701.1 (6 %)
prefill tok/s, 8k document	1062.5	1419.7	+34 %	10484.7	14 %	739.6 (7 %)
prefill tok/s, 16k document	887.4	1337.8	+51 %	9014.9	15 %	702.9 (8 %)
prefill tok/s, 32k document	632.6	1176.6	+86 %	7040.8	17 %	671.6 (10 %)
4 streams, aggregate decode tok/s	125.6	138.2	+10 %	220.5	63 %	115.7 (52 %)
speculative decoding, drafted tokens accepted	—	68 % ¹ †	—	n/a	n/a	—

• llama.cpp: build llamacpp-rocm/b1355, server arguments -ngl 99 -cram 0 --spec-type draft-dflash --spec-draft-model ~/.cache/huggingface/hub/models--incoai--Flint-DFlash2-GGUF/snapshots/74e6ac0c538cfd393d97bd93694e5c4c6b6aa8c3/Flint-DFlash2-Q4_K_M.gguf --spec-draft-n-max 3 -ngld 99 -c 65536 -kvu -np 4.
• Loom: speculative decoding on (DFlash, depth 3).
• Same GGUF on both engines, tokens per second. Loom: client-side timing; llama.cpp: server-side timings. Runs per value: decode and prefill 2 (mean), 4 streams 3 (median).
• Measured 2026-11-06; device AMD RYZEN AI MAX+ 395 w/ Radeon 8060S, AMD Radeon 8060S Graphics, Ubuntu 26.04; hrx-loom 0.1.0 (hrx-system d01d1c1); source: local run.
• ¹ Measured 2026-11-06; device AMD RYZEN AI MAX+ 395 w/ Radeon 8060S, AMD Radeon 8060S Graphics, Ubuntu 26.04; hrx-loom 0.1.0 (hrx-system d01d1c1, recovered); source: local run.
• † Measured on programs other than this repository's (sha256 8cb17fa, this repository's 5422915): the value describes another build of the package until a session re-measures it.
• Drafted tokens accepted: the share of the drafter's proposed tokens the model accepted during the spec gate's requests.
• Roofline: the hardware ceiling for this GGUF on AMD Ryzen AI Max+ 395 "Strix Halo" (gfx1151) at 216 GB/s and 59 TFLOP/s, from loom-roofline roofline.py (loom-skills 4da7fb0); decode and 4 streams with speculative decoding: verify passes × (1 + acceptance × depth 3) tokens per pass, the drafter's cost not counted; acceptance measured in the row's own session: decode 49 %, decode at 2k 52 %, decode at 32k 38 %, 4 streams 41 %. DevLab: Loom alone on an AMD DevLab Dispatch runner (DevLab Dispatch run 42315858300, runner X2-LA03-S04-D07, 2026-11-07); compare it with the Loom column, not with llama.cpp.
• — not measured; n/a not applicable.

Features

feature	status	detail
API	✅	OpenAI-style /v1/chat/completions
tool calling	✅	OpenAI-style tools and tool_calls
reasoning / thinking	✅	always on
reasoning effort	❌	not supported
speculative decoding	✅	DFlash, depth 3, on by default; drafter incoai/Flint-DFlash2-GGUF
image input	❌	not supported
audio input / output	❌	not supported
sampling	✅	temperature, top_p, top_k, seed
context size	✅	262,144 tokens per request
concurrency	✅	4 concurrent requests
max output tokens	✅	4,096 (preset max_tokens)

Correctness

gate	description	result
kernel_tests	Unit tests of the generated kernels against reference implementations	❌ no evidence
golden	llama.cpp reference recorded on the same GGUF: greedy tokens and top-5 log-probs, short prompts and 1k-32k documents	✅ pass ¹
compare	Greedy output vs the llama.cpp reference, token by token; every divergence is a near-tie	✅ pass ¹
vision	Image input vs llama.cpp with the same projector: identical prompt tokens, replies equal up to near-ties	— not applicable
consistency	Concurrent requests give the same output as the same requests run alone	✅ pass ¹
spec	Speculative decoding on gives exactly the output of speculative decoding off	✅ pass ¹
agent	Multi-turn tool-calling loop: exact prompt rendering, history reuse, valid tool calls, output vs llama.cpp	✅ pass ¹
agent_think_off	The tool-calling loop with thinking off	— not applicable
sampling	temperature, top_k and seed behave as specified; first-token distribution matches llama.cpp	✅ pass ¹
needle	Needle-in-a-haystack prompt at up to half the context, compared with llama.cpp; over-capacity prompt refused	✅ pass ²
needle_full	Needle-in-a-haystack prompt at the full context is served	✅ pass ¹
decision	Decision answers (choice, score, noul probabilities) vs llama.cpp on the same prompts	— not applicable

• ¹ DevLab Dispatch run 42315858300, LEVEL=full, 2026-11-07; device AMD RYZEN AI MAX+ 395 w/ Radeon 8060S, AMD Radeon 8060S Graphics, Ubuntu 24.04; hrx-loom 0.1.0 (hrx-system d01d1c1).
• ² DevLab Dispatch run 42285876725, LEVEL=dev, 2026-11-06; device not recorded; hrx-loom 0.1.0 (hrx-system d01d1c1, recovered).
• Gates are the loom-benchmark correctness gates (gates.sh); a gate without a recorded run shows no evidence.

Files

• model.json / model.json.lock: the card and the pinned source files (sha256 checked by loom-fetch).
• program/*.loom, weights.loom: the generated Loom programs.
• tokenizer.json: the tokenizer.
• llm_config.json: the configuration the programs were generated from.
• info.json: the measurements, gate evidence and versions this card shows; everything else comes from the files above.
• REPORT.md: how the values above were measured: setup, commands and the value of every run.

Versions: card schema loom-card/2; loom-skills af9f2a0 (recovered); hrx-loom 0.1.0 (hrx-system d01d1c1, recovered); card generator 4da7fb0.

Source files, fetched by loom-fetch from their own repositories:

• checkpoint: ggml-org/Flint-GGUF @ 468307a, Flint-Q4_K_M.gguf
• drafter: incoai/Flint-DFlash2-GGUF @ 74e6ac0, Flint-DFlash2-Q4_K_M.gguf