Experiment Complete Published
Ollama 0.40.13 on M5 Max · 64 GiB, Metal
LFM2.5 350M as Q4_K_M (4-bit), archive 256 input / 128 output: a 256-token prompt and 128 tokens out. Measured 12 Nov 2026, 20:38 UTC by Glenn Sonna, Xybrid.
Eligible for ranking when the objective's required metrics are measured and this configuration is selected.Approximate comparison: The experiments use different effective KV cache precision: bf16, fp32, litert-lm default, f16, magnitude default, model-dtype, uzu default.
Experiment id s3-lfm2.5-350m-4-bit-ollama-256
Summary
Over 6 valid trials of 6 attempted (6 planned), computed from the raw trials below.
| Metric | Median | Range | Mean ± SD | Quartiles |
|---|---|---|---|---|
| Decode speed tok/s, higher is better | 742.8 | 734.2–745.7 | 741.0 ± 5.2 | 736.7–745.1 |
| Time to first token ms, lower is better | 21.4 | 21.3–21.6 | 21.5 ± 0.1 | 21.4–21.5 |
| Peak memory GiB, lower is better | 0.46 | 0.46–0.47 | 0.46 ± 0.00 | 0.46–0.46 |
| Prefill speed tok/s, higher is better | 12,795 | 12,754–12,870 | 12,800 ± 43.2 | 12,770–12,819 |
Trials, metric by metric
The line runs from the lowest to the highest valid trial, the dot is the median and each tick is one trial.
Every trial
Raw token counts and durations, with the speeds computed from them. Invalid trials stay in the record with their reason; they are left out of the summary, never silently dropped.
| Trial | Time to first token ms | Prefill tok/s | Decode tok/s | Peak memory GiB | Thermal | Note |
|---|---|---|---|---|---|---|
| 1 | 21.5 | 12,769256 tok in 20.0 ms, engine | 745.7127 tok in 170 ms, engine | 0.47 | nominal → nominal | |
| 2 | 21.4 | 12,815256 tok in 20.0 ms, engine | 744.9127 tok in 171 ms, engine | 0.46 | nominal → nominal | |
| 3 | 21.6 | 12,774256 tok in 20.0 ms, engine | 745.2127 tok in 170 ms, engine | 0.46 | nominal → nominal | |
| 4 | 21.5 | 12,821256 tok in 20.0 ms, engine | 735.3127 tok in 173 ms, engine | 0.46 | nominal → nominal | |
| 5 | 21.4 | 12,754256 tok in 20.1 ms, engine | 734.2127 tok in 173 ms, engine | 0.46 | nominal → nominal | |
| 6 | 21.3 | 12,870256 tok in 19.9 ms, engine | 740.8127 tok in 171 ms, engine | 0.46 | nominal → nominal |
Caveats
- Publisher cards identify the base model but do not always pin the exact export source checkpoint revision.
- Engines may differ in quantization scheme, effective KV cache precision, kernels, and activations.
Configuration
Device
- Device
- Apple Mac17,6
- Cohort
- Apple silicon Macs (desktop)
- SoC
- Apple M5 Max
- CPU
- 18-core Apple CPU
- GPU
- Not recorded
- NPU
- Not recorded
- RAM
- 64.00 GiB
- OS
- macOS 27.2.0 (26B123)
- Kernel
- 27.2.0
- Power
- mains
- Low power mode
- No
- Performance mode
- Not recorded
Engine
- Engine
- Ollama
- Maintainer
- Ollama
- License
- MIT, open source
- Version
- 0.40.13
- Commit
- Not recorded
- Binary SHA-256
- 270089e59e1d411ac4d0987154b4e95f48b29a81c9d274833c6b766d6df88300
- Built from
- release
- Build flags
- None
- Backend
- Metal
- Threads
- Not recorded
- Interface
- http-server
- Settings in effect
- backend=metal model=xybrid-7e6f72643caafc9a blob=sha256-7e6f72643caafc9a68256686638c4d7916f2cec76d1df478d4c3ddcd95a6aed4 offloaded_layers=17 bos_stripped=true llama_server.host=127.0.0.1 llama_server.no-webui=true llama_server.offline=true llama_server.c=4096 llama_server.np=1 llama_server.log-verbosity=4 llama_server.no-log-prefix=true llama_server.no-log-timestamps=true llama_server.cache-type-k=f16 llama_server.cache-type-v=f16 llama_server.flash-attn=on llama_server.b=512 llama_server.ub=512 llama_server.context-shift=true llama_server.keep=4 env.OLLAMA_KEEP_ALIVE=-1 env.OLLAMA_FLASH_ATTENTION=1 env.OLLAMA_KV_CACHE_TYPE=f16 env.OLLAMA_NOHISTORY=1 env.OLLAMA_NUM_PARALLEL=1 env.OLLAMA_MAX_LOADED_MODELS=1 env.LLAMA_ARG_CACHE_RAM=0
Model and weights
- Base model
- LFM2.5 350M (LFM2.5, 350M)
- Upstream
- LiquidAI/LFM2.5-350M@9921cd4d82c0e3986887aa4b376f9446729ef8a3
- Weights
- LiquidAI/LFM2.5-350M-GGUF@657e078c94084481950a2d555a941481f715536b
- Format
- gguf
- File 1
- LFM2.5-350M-Q4_K_M.gguf · 218.7 MiB · sha256 7e6f72643caafc9a68256686638c4d7916f2cec76d1df478d4c3ddcd95a6aed4
- File 2
- README.md · 3,277 bytes · sha256 4773739eb49540664b16566dc655d03a51df3937639e9dbf28d48a15d01b2436
- Quantization
- Q4_K_M (4-bit)
- Weight bits
- 4
- Group size
- Not recorded
- KV cache
- f16
- Conversion
- Published upstream by LiquidAI
- Conversion command
- Not recorded
Workload
- Workload
- Archive 256 input / 128 output
- Prompt
- 256 tokens, from prompts/source.txt
- Prompt SHA-256
- c10d1b7bef62d95322f8f8d229dc66118d9375107bdc49b06f76287d058b2f6c
- Chat template
- ba551d58630afa3190b1be3602e28301f3d2e9bbac978dfc49d6d825171648b6
- Output
- 128 tokens, ignoring end of sequence
- Context length
- 4,096
- Batch size
- 1
- Sampling
- Greedy
- Timing
- One streamed request per trial
- Warmup
- 2 runs, not counted
- Trials planned
- 6
- Cooldown
- 15 s between trials
Conditions
- Ambient
- Not recorded
- Thermal at start
- nominal
- Thermal at end
- nominal
- Thermal gate
- macOS thermal state nominal
- Throttling detected
- No
Provenance
- Measured by
- Glenn Sonna, Xybrid
- Operator affiliation
- Xybrid
- Run at
- 12 Nov 2026, 20:38 UTC
- Harness
- xybrid-bench 0.4.0 (f67b01e5a1a77cc89f5984f23e35dd4590fa1bec)
- Methodology
- v0.6
- Publication
- published
- Published
- 16 Nov 2026, 10:23 UTC
- Reviewed by
- Glenn Sonna
Output check
- Result
- Passed
- Method
- sanity-v1-bf16-reference
- Reference
- checks/sanity-v1.json
- Reference SHA-256
- a702dee079525ae0ec90465b1c9e26052c99cb3e1c53e83dea05be4c194fc7ca
- Required passing checks
- 7
- Check arithmetic
- Failed · sha256 9f14025af0065b30e47e23ebb3b491d39ae8ed17d33739e5ff3827ffb3634953
- Check capital
- Passed · sha256 5dd272b4f316b776a7b8e3d0894b37e1e42be3d5d3b204b8a5836cc50597a6b1
- Check opposite
- Passed · sha256 f151cd9e12c8994edfa316e4624f956bc68460454869177f287146f8cc956a12
- Check weekday
- Passed · sha256 7d8af1de1262f150187d5938bf649d6c35726970685f916d1ec6392a801b9762
- Check color
- Passed · sha256 d486dfbd5fb578340bccbdd0a662527eab38550648d5f44517e5ac71b8824703
- Check sort
- Failed · sha256 edb794d615a63f5eec436d83bed7299d6af36f508c0bf38a3c1cd85d6e4ebd56
- Check animal
- Passed · sha256 4b227777d4dd1fc61c6f884f48641d02b4d121d3fd328cb08b5531fcacdabf8a
- Check word
- Failed · sha256 639807a2e01886a2236b9fe1fe9bdbdea267b9c84c4acb3ab0858a6a13531b9a
- Check copy
- Passed · sha256 d61e6890d0e3571b0c7693612d41f8d059a1fcdef4ba959c0af4144541d41907
- Check needle-start
- Passed · sha256 8b9af2e5b8b4cfddb87d171997b95c471f8d68a68e30fb5177abeb3332372edd
- Check needle-end
- Passed · sha256 e6589c736af32f81bf239c5eda6de41959122ccc1d05463b75e0eb0c70fc831c
- Check continuation-128
- Failed · sha256 ce02aec498756856567446a2f4cb35b219ea726d81089769eddd1499c99ba95f
- Detail
- 8/8 of the checks the model passes at BF16; needs 7 and the mandatory ones (needle-start, needle-end). Left out, failed by every BF16 configuration: arithmetic, sort, word, continuation-128.
Reproduce it
pnpm bench run --plan plans/s3-lfm2.5-350m-4-bit.json
Three palindromic blocks per quantization class and workload; load, two warmups, one trial, unload per turn.
Raw data
- This record, as committed JSON, schema v0.3.0
- Complete finalized session evidence raw ·
sha256 53feb733252e…