Repository navigation
results: Nanbeige4.3-33B on Arc Pro B70 (llama.cpp q4_k_m, vLLM int4 TP2, ngram) - #118
Merged
Merged
Conversation
…TP2, ngram) Three runs measured 2026-10-28 on a 2x Arc Pro B70 rig (rig 1), one stream each: - llama.cpp SYCL, Q4_K_M, one card: 23.48 tok/s decode (tg128), 1083.8 tok/s pp8192 - vLLM-XPU TP2, W4A16 int4, fp8 KV: 41.01 tok/s decode, 1420 tok/s prompt, 262144 ctx - same TP2 route with ngram speculation at a 98K prompt: 30.31 tok/s (+27 %) Stacked on the catalog PR: the model id and its int4/q4_k_m boards must be deployed before ingestion can pass.
|
@SergiioB is attempting to deploy a commit to the Community Labs Team on Vercel. A member of the Team first needs to authorize it. |
- results/SergiioB/2026-10-28-nanbeige4-3-33b-int4-vllm-tp2-ngram.json: imported (https://intelinside.ai/results/284) - results/SergiioB/2026-10-28-nanbeige4-3-33b-int4-vllm-tp2.json: imported (https://intelinside.ai/results/285) - results/SergiioB/2026-10-28-nanbeige4-3-33b-q4-k-m-llamacpp.json: imported (https://intelinside.ai/results/286) Merge ingestion completed. Rerunning this workflow will not duplicate these submissions. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Three measured runs on rig 1 (2× Arc Pro B70), uploaded as files.
...q4-k-m-llamacpp.json1692f9e50...int4-vllm-tp2.json...int4-vllm-tp2-ngram.jsonAll three validate with
node frontend/scripts/validate-results.mjs.llama-bench, 3 reps); vLLM cells are client streamed TTFT/decode.evidenceUrlpoints at the cookbook recipe with commands, artifacts, hashes and the untested list. Quality on the same build: 40/40 on a 40-item deterministic suite.Stacked on #117 (catalog row for the model). Ingestion will fail its catalog check until that is merged and the catalog is deployed; happy to rebase or rerun when it is.