loom-fun/LiquidAI-d1-2B-loom

d1-2B (Q4_K_M) for AMD Loom

A Loom package (.loom programs and a model.json card) for the unmodified GGUF checkpoint LiquidAI/d1-2B-GGUF (d1-2B-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/LiquidAI-d1-2B-loom/resolve/main/model.json --materialize=copy --output=d1-2B-GGUF
loom-serve --model=d1-2B-GGUF/model.json --device=amdgpu

The server listens on 127.0.0.1:8080 and answers SystemOne decision requests at /v1/systemone.

Performance vs llama.cpp

metric	llama.cpp	Loom	margin
decode tok/s, short prompt	n/a	n/a	n/a
decode tok/s, 2k context	n/a	n/a	n/a
decode tok/s, 32k context	n/a	n/a	n/a
prefill tok/s, 1k document	5926.4	7041.3	+19 %
prefill tok/s, 2k document	5684.0	7138.1	+26 %
prefill tok/s, 8k document	4204.3	5488.4	+31 %
prefill tok/s, 16k document	3128.7	4637.4	+48 %
prefill tok/s, 32k document	—	—	—
4 streams, aggregate decode tok/s	n/a	n/a	n/a

• llama.cpp: ROCm build b1357 (lemonade-sdk/llamacpp-rocm), speculative decoding off.
• Loom: speculative decoding off.
• Same GGUF on both engines, tokens per second. Loom: client-side timing; llama.cpp: server-side timings. Runs per value: decode and prefill not recorded (mean), 4 streams not recorded (median).
• Measured 2026-10-29; source: loom-factory run 40842422037.
• — not measured; n/a not applicable.

Features

feature	status	detail
API	✅	/v1/systemone: choice, score and noul decisions
tool calling	—	not applicable
reasoning / thinking	—	not applicable
reasoning effort	—	not applicable
speculative decoding	—	not applicable
image input	—	not applicable
audio input / output	❌	not supported
sampling	—	not applicable
context size	❌	not verified
concurrency	✅	64 pending requests
max output tokens	—	not applicable

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	— not applicable
compare	Greedy output vs the llama.cpp reference, token by token; every divergence is a near-tie	— not applicable
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	— not applicable
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	— not applicable
needle	Needle-in-a-haystack prompt at up to half the context, compared with llama.cpp; over-capacity prompt refused	— not applicable
needle_full	Needle-in-a-haystack prompt at the full context is served	— not applicable
decision	Decision answers (choice, score, noul probabilities) vs llama.cpp on the same prompts	✅ pass ¹

• ¹ loom-factory run 40842422037, 2026-10-29.
• 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 data this card is rendered from (schema loom-card/1).

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

• checkpoint: LiquidAI/d1-2B-GGUF @ e968f65, d1-2B-Q4_K_M.gguf