Hackernews posts about 70B

1. Show HN: Llama 3.1 70B on a single RTX 3090 via NVMe-to-GPU bypassing the CPU (github.com)
   395 points by xaskasdf 29 days ago | 101 comments
2. QuarterBit – Train 70B LLMs on a single GPU (quarterbit.dev)
   3 points by quarterbit 19 days ago | 4 comments
3. Testing Nvidia's FP4: Running 70B LLMs on a Single RTX 5090 with Real Benchmarks (ai.gopubby.com)
   2 points by Aedelon 12 days ago | discuss
4. TinyTTS: Ultra-light English TTS (9M params, 20MB), 8x CPU, 67x GPU
   5 points by letrghieu 26 days ago | discuss
5. Show HN: I built a LLM human rights evaluator for HN (content vs. site behavior) (observatory.unratified.org)
   3 points by 9wzYQbTYsAIc 19 days ago | 2 comments
6. Show HN: ContextFlow – YouTube videos to viral social content (built by a 15yo) (contextflowai.online)
   3 points by Terminator_XD 17 days ago | discuss
7. Show HN: HushBrief – A stateless, zero-retention AI document summarizer (hushbrief.app)
   3 points by FidelitasLLC 21 days ago | discuss
8. Show HN: SayIt – AI that helps you prepare for difficult conversations (www.sayit.chat)
   2 points by DivakaranK 15 days ago | 2 comments
9. Show HN: Sleeping LLM – A language model that remembers by sleeping (github.com)
   2 points by vbaranov87 25 days ago | discuss
10. A small tool I made for local LLMs: LLM-neofetch-plus
    1 points by HFerrahoglu 26 days ago | 2 comments
11. Show HN: QLoRA fine-tuning in .zse INT4 format by ZSE
    1 points by zyoralabs 19 days ago | discuss
12. Show HN: SLOK – SLO composition with traffic-weighted service chains in K8s
    1 points by lep_qq 19 days ago | discuss
13. Show HN: We attacked 8 AI agents with adversarial probes – 5 failed (antarraksha.ai)
    1 points by akash1039ai 19 days ago | discuss
14. Show HN: TechGrill – AI Interview Practice (techgrill.vercel.app)
    1 points by wasivis 25 days ago | discuss
15. Binance fired employees who found $1.7B in crypto was sent to Iran (www.nytimes.com)
    552 points by boplicity 28 days ago | 265 comments
16. Nvidia PersonaPlex 7B on Apple Silicon: Full-Duplex Speech-to-Speech in Swift (blog.ivan.digital)
    374 points by ipotapov 18 days ago | 125 comments
17. Britain is ejecting hereditary nobles from Parliament after 700 years (apnews.com)
    310 points by divbzero 11 days ago | 397 comments
18. 70k Books Found in Hidden Library in This Germany Home (2023) (bookstr.com)
    118 points by eatonphil 17 days ago | 79 comments
19. Open Source Iran War Cost Tracker: 45.7B (iranwarcost.com)
    50 points by koverda 19 days ago | 20 comments
20. We may soon have 70M boomers too old to drive, too car-dependent to stop (lloydalter.substack.com)
    36 points by cainxinth 20 days ago | 11 comments
21. The 1,700-year-old megastructure history almost forgot (www.cnn.com)
    28 points by simonebrunozzi 21 days ago | 3 comments
22. Compress Your Claude.md: Cut 60-70% of System Prompt Bloat in Claude Code (techloom.it)
    25 points by jchilcher 26 days ago | 9 comments
23. Russia suffers deadliest day of year with 1,700 troops lost (www.telegraph.co.uk)
    20 points by doener 3 days ago | 4 comments
24. 70% of Australian teens are getting around the social media ban (twitter.com)
    15 points by delichon 14 days ago | 8 comments
25. We Have 700 AI Agents Playing a Game We Don't Understand (www.spacemolt.com)
    14 points by statico 3 days ago | 1 comments
26. Wolbachia-infected mosquitoes suppress dengue by 70% across Singapore (medicalxpress.com)
    13 points by marojejian 28 days ago | discuss
27. I emailed 70 consulting partners. No replies. What it taught me (willem720055.substack.com)
    12 points by anqer 12 days ago | 12 comments
28. Five People in Their 60s, 70s, and 80s Share How They Plan to Age at Home (www.nytimes.com)
    12 points by bookofjoe 20 days ago | 2 comments
29. Covenant-72B is the largest decentralized LLM pre-training run in history (twitter.com)
    11 points by _____k 2 days ago | discuss
30. I used Claude to match 200 Clinical Trials to 700 PubMed Papers (everyrow.io)
    10 points by peter94 28 days ago | 2 comments