tzhouamandClaude Opus 5
9264ff81ab
The packaging commit shipped the code and a summary of the numbers. It left out
everything the numbers are measured *against*, which is most of what makes them
checkable.
Added:
baselines/ llama.cpp runs (thread sweeps, NUMA sweeps, the
interleaved A/B passes behind the headline decode
figure) and the quality references: each build
against its own bf16 over 12 greedy prompts.
benchmarks/.../results/ 102 artifacts backing the published figures, each
with a provenance block naming the model,
quantization, thread set and build SHAs.
docs/engines/ the engine analyses -- llama.cpp, ExecuTorch, MNN,
ncnn, vLLM, vLLM-Omni -- and the cross-engine and
device-simulation comparisons.
docs/analysis/ the full study documents the summaries distil.
docs/edge/comparisons.md where this fork sits against llama.cpp and why:
decode is llama.cpp's, prefill is ours by ~2x, and
since neither engine exceeds 39% of the measured
222 GB/s ceiling the decode gap is the execution
model rather than the kernels.
docs/edge/environment.md pins and machine, and the correction below.
On "the modified vllm": there isn't one, and the doc says so plainly because it
is the obvious assumption. The vLLM checkout is clean upstream (0 commits
ahead, 0 uncommitted) and the installed wheel is stock 0.28.0+cpu. This fork
changes vLLM's CPU *behaviour* through vLLM's own extension points at import
time -- three CPU patches in vllm_omni/patch.py, the quantization-config
registry, the IR op registry, platform hooks -- specifically so the changes
survive a wheel upgrade rather than being lost on it. One of those patches is
why "KV sizing is not a memory lever" looked true for months: CpuPlatform
overwrote kv_cache_memory_bytes unconditionally, so the setting had no effect.
There is no patch series to apply; pin the version and let patch.py do the rest.
.gitignore gains explicit exceptions: upstream's blanket `*.json` and `results/`
rules were silently eating the measurement artifacts, which in this fork are the
substance rather than build output. Binary captures stay ignored.
No vLLM or llama.cpp source is redistributed -- only output produced with them.
NOTICE records both.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L1ax5wYLA6XbqEL1i8L9gZ
50 lines
4.4 KiB
JSON
50 lines
4.4 KiB
JSON
{
|
|
"chat_short": {
|
|
"text": "We are asked: \"What is the capital of France? Answer in one word.\" So the answer should be a single word: \"Paris\". But we need to output just the answer in one word. The instruction says: \"Answer in one word.\" So we should respond with \"Paris\".</think>Paris",
|
|
"token_ids": null
|
|
},
|
|
"chat_long": {
|
|
"text": "We are given an inventory listing with items 1 through 120. Each item has a code for a warehouse district. The codes are given as:\n\nItem 1: 1007\nItem 2: 1014\nItem 3: 1021\n",
|
|
"token_ids": null
|
|
},
|
|
"chat_tools": {
|
|
"text": "The user wants weather for Shenzhen in celsius. I have the weather function available. Let me call it with the city \"Shenzhen\" and unit \"celsius\".</think><tool_call>get_weather<arg_key>city</arg_key><arg_value>Shenzhen</arg_value><arg_key>unit</arg_key><arg_value>celsius</arg_value></tool_call>",
|
|
"token_ids": null
|
|
},
|
|
"plain_0": {
|
|
"text": " the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia is the capital of Australia",
|
|
"token_ids": null
|
|
},
|
|
"plain_1": {
|
|
"text": " mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission mission",
|
|
"token_ids": null
|
|
},
|
|
"plain_2": {
|
|
"text": " a, b = 0, 1\n for _ in range(n):\n a, b = b, a + b\n return b\n\n\nfib = fibonacci\n\n\nprint(fib(10))\nprint(fib(20))\nprint(fib",
|
|
"token_ids": null
|
|
},
|
|
"plain_3": {
|
|
"text": " that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that that",
|
|
"token_ids": null
|
|
},
|
|
"plain_4": {
|
|
"text": " to make a proper espresso you need to make a proper espresso you need to make a proper espresso you need to make a proper espresso you need to make a proper espresso you need to make a proper espresso you need to make a proper espresso you need to make a proper espresso you need to make a proper espresso you need to",
|
|
"token_ids": null
|
|
},
|
|
"plain_5": {
|
|
"text": " the difference between TCP and UDP is the difference between TCP and UDP is the difference between TCP and UDP is the difference between TCP and UDP is the difference between TCP and UDP is the difference between TCP and UDP is the difference between TCP and UDP is the difference between TCP and UDP is the difference between TCP and UDP is the",
|
|
"token_ids": null
|
|
},
|
|
"plain_6": {
|
|
"text": " converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts converts",
|
|
"token_ids": null
|
|
},
|
|
"plain_7": {
|
|
"text": " is used to organize data in a way that is used to organize data in a way that is used to organize data in a way that is used to organize data in a way that is used to organize data in a way that is used to organize data in a way that is used to organize data in a way that is",
|
|
"token_ids": null
|
|
},
|
|
"plain_8": {
|
|
"text": " 1621. The Treaty of Westphalia was signed in 1621. The Treaty of Westphalia was signed in 1621. The Treaty of Westphalia was signed in 1621. The Treaty of Westphalia was signed in 162",
|
|
"token_ids": null
|
|
}
|
|
} |