loom-fun/rippertnt-HyperCLOVAX-SEED-Text-Instruct-1.5B-loom

HyperCLOVAX-SEED-Text-Instruct-1.5B-Q4_K_M for AMD Loom

A Loom package (.loom programs and a model.json card) for the unmodified GGUF checkpoint rippertnt/HyperCLOVAX-SEED-Text-Instruct-1.5B-Q4_K_M-GGUF (hyperclovax-seed-text-instruct-1.5b-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/rippertnt-HyperCLOVAX-SEED-Text-Instruct-1.5B-loom/resolve/main/model.json --materialize=copy --output=HyperCLOVAX-SEED-Text-Instruct-1.5B
loom-serve --model=HyperCLOVAX-SEED-Text-Instruct-1.5B/model.json --device=amdgpu --memory_bytes=24000000000

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	DevLab llama.cpp	DevLab Loom	local ÷ DevLab (llama.cpp, Loom)
decode tok/s, short prompt	99.4	171.7	+73 %	230.3	75 %	128.1	151.8 ¹	0.78, 1.13
decode tok/s, 2k context	99.0	201.9	+104 %	307.7	66 %	100.6	124.1 ¹	0.98, 1.63
decode tok/s, 32k context	93.3	112.8	+21 %	145.3	78 %	—	108.2 ¹	—, 1.04
prefill tok/s, 1k document	4067.8	6840.4	+68 %	20950.1	33 %	4753.4	6378.0 ¹	0.86, 1.07
prefill tok/s, 2k document	3633.8	6944.0	+91 %	20243.5	34 %	4081.8	6406.1 ¹	0.89, 1.08
prefill tok/s, 8k document	3115.4	6141.4	+97 %	16836.3	36 %	2233.4	4636.9 ¹	1.39, 1.32
prefill tok/s, 16k document	2020.9	4930.7	+144 %	13750.4	36 %	1597.2	4003.1 ¹	1.27, 1.23
prefill tok/s, 32k document	1334.0	3538.1	+165 %	10062.0	35 %	1037.8	3052.2 ¹	1.29, 1.16
4 streams, aggregate decode tok/s	171.4	447.2	+161 %	802.3	56 %	441.3	359.8 ¹	0.39, 1.24
speculative decoding, drafted tokens accepted	—	7 % ²	—	n/a	n/a	n/a	n/a	n/a

• llama.cpp: build llamacpp-rocm/b1369, server arguments -ngl 99 -cram 0 --spec-type ngram-simple -c 35024 -kvu -np 4.
• Loom: speculative decoding on (n-gram prompt lookup, 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 2026-11-08; 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 2026-11-08; device not recorded; hrx-loom unknown; source: DevLab Dispatch run 42498519037.
• ² Measured 2026-11-08; device AMD RYZEN AI MAX+ 395 w/ Radeon 8060S, AMD Radeon 8060S Graphics, Ubuntu 24.04; hrx-loom 0.1.0 (hrx-system d01d1c1); source: DevLab Dispatch run 42501241494.
• DevLab: the same measurement on an AMD DevLab Dispatch Strix Halo runner; local ÷ DevLab is the local value over the DevLab value of each engine, recorded for comparison, not a verdict.
• 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 ecb661b); 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 2 %, decode at 2k 24 %, decode at 32k 60 %, 4 streams 5 %.
• — not measured; n/a not applicable.

Features

feature	status	detail
API	✅	OpenAI-style /v1/chat/completions
tool calling	❌	not supported
reasoning / thinking	❌	not supported
reasoning effort	❌	not supported
speculative decoding	✅	n-gram prompt lookup, depth 3, on by default
image input	❌	not supported
audio input / output	❌	not supported
sampling	✅	temperature, top_p, top_k, seed
context size	✅	131,072 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	✅ pass (37/37) ¹
golden	llama.cpp reference recorded on the same GGUF: greedy tokens and top-5 log-probs, short prompts and 1k-32k documents	✅ pass ²
compare	Greedy output vs the llama.cpp reference, token by token; every divergence is a near-tie	✅ pass ²
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 ²
spec	Speculative decoding on gives exactly the output of speculative decoding off	✅ pass ²
agent	Multi-turn tool-calling loop: exact prompt rendering, history reuse, valid tool calls, output vs llama.cpp	— not applicable
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 ²
needle	Needle-in-a-haystack prompt at up to half the context, compared with llama.cpp; over-capacity prompt refused	✅ pass ³
needle_full	Needle-in-a-haystack prompt at the full context is served	✅ pass ²
decision	Decision answers (choice, score, noul probabilities) vs llama.cpp on the same prompts	— not applicable

• ¹ DevLab Dispatch run 42332379039, 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 42501241494, LEVEL=full, 2026-11-08; device AMD RYZEN AI MAX+ 395 w/ Radeon 8060S, AMD Radeon 8060S Graphics, Ubuntu 24.04; hrx-loom 0.1.0 (hrx-system d01d1c1).
• ³ DevLab Dispatch run 42332379039, 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).
• 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 e7cb8bc; hrx-loom 0.1.0 (hrx-system d01d1c1); card generator ecb661b.

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

• checkpoint: rippertnt/HyperCLOVAX-SEED-Text-Instruct-1.5B-Q4_K_M-GGUF @ 3d2edd5, hyperclovax-seed-text-instruct-1.5b-q4_k_m.gguf