Running 35B–400B LLMs on a GPU-less Cluster to Mine 10,000 Papers — and the 4 Bugs That Almost Ruined the Data

Jun 3, 2026 · 5:55 AM UTC ·10 min read · 0 reactions · 0 comments · 42 views

#machinelearning #dataextraction #infrastructure

TL;DR · WeSearch summary

A team successfully built a CPU-only distributed LLM pipeline to extract structured data from 10,000 research papers. The project faced challenges, particularly with data quality, as four significant bugs were discovered during the process. The architecture utilized open-source tools and demonstrated that effective LLM extraction is possible without GPUs, focusing on correctness over speed.

Key facts

▪The team operated an internal research cluster with older x86 servers and no GPUs.
▪They aimed to extract structured data for meta-analysis from approximately 10,000 full-text research papers.
▪The architecture included a MoE model and a vector database, relying solely on CPU resources.

About this source

DEV.to (Top) files mainly under programming. We currently carry 4,924 of its stories.

All DEV.to (Top) coverage →

Original article

DEV.to (Top)

Read full at DEV.to (Top) →

Story provenance

Source · retrieval · rights · ranking — open for full record

inspect →

Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.

Record

Original publisher	DEV.to (Top)
Canonical URL	https://dev.to/sysoft/running-35b-400b-llms-on-a-gpu-less-cluster-to-mine-10000-papers-and-the-4-bugs-that-almost-ka3
Publication time	Wed, 03 Jun 2026 05:55:34 +0000
Retrieval time	2026-06-03T06:11:56.795Z
Last seen	2026-06-03T06:11:56.795Z
Headline source	Publisher (no WeSearch rewrite)
Excerpt source	publisher body
Excerpt method	First ~120 words (~800 chars) of extracted publisher body, fair-use limited.
Summary	WeSearch · cerebras-chat (WeSearch summarizer)
Summary source text	contentText
Citation coverage	Summary is a WeSearch-generated derivative; primary citation is the original publisher URL.
Cluster	dJO2566L4j13
Cluster logic	Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison.
Ranking reason	Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking.
Publisher visit	Yes — open original
Substitutes article?	No — link-out required for full text

Rights status (four layers)

Publisher-declared

No publisher-confirmed rights record for this source yet.

Machine-readable

No source-specific machine-readable restriction detected beyond the public feed.

WeSearch interpretation

WeSearch declared handling (basis: Derived from the published RSS/Atom feed). This is WeSearch policy, not a legal grant on the publisher's behalf.

Unknown

Retrieval and training permissions are not asserted unless the publisher confirms them.

WeSearch handling by dimension

Indexing	May the item be indexed (stored, ranked, made findable)?	Allowed
Snippet	May a short excerpt of the publisher's text be shown?	Allowed
AI summary	May WeSearch generate its own short summary of the article?	Limited
Retrieval / RAG	May the content be exposed for third-party retrieval-augmented generation?	Not asserted
Model training	May the content be used to train AI models?	Not asserted
Commercial reuse	May the content be reused commercially?	Not permitted

Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.

Open original publisher → Source rights JSON → Machine-readable provenance → Rights framework → WESEARCH-STORY-PROVENANCE-V1

Opening excerpt (first ~120 words) tap to expand

try { if(localStorage) { let currentUser = localStorage.getItem('current_user'); if (currentUser) { currentUser = JSON.parse(currentUser); if (currentUser.id === 3962195) { document.getElementById('article-show-container').classList.add('current-user-is-article-author'); } } } } catch (e) { console.error(e); } byeongsoo kang Posted on Jun 3 • Originally published at bric.pe.kr Running 35B–400B LLMs on a GPU-less Cluster to Mine 10,000 Papers — and the 4 Bugs That Almost Ruined the Data #llm #machinelearning #python #infrastructure A field report from building a CPU-only, distributed LLM pipeline for large-scale scientific literature extraction. No GPUs. A lot of quantization. And four silent data-quality bugs that taught me more than the happy path ever did.

…

Excerpt limited to ~120 words for fair-use compliance. The full article is at DEV.to (Top).

Anonymous · no account needed

Discussion

0 comments