Top all-time · #15366
Educating Silicon
Personal blog · educatingsilicon.comhardwareai
Educating Silicon is a personal blog at educatingsilicon.com about hardware and ai. Since 2016, 5 different Hacker News readers have linked it 7 times, most often “How much LLM training data is there from all sources in Trillions of Tokens?”. No feed was found, so read it on the site itself.
Too few sampled comments (1) to say how often HN recommends it.?
Why HN links it
HN readers link this blog for a paper on Google's OCR system compared against commercial benchmarks. The only comment cites it as a source on OCR accuracy for text in photographs.
Summarised by Claude from a sample of 1 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Most-linked posts?
- How much LLM training data is there from all sources in Trillions of Tokens?3 people · 3 mentions · first linked 2024-05-10
- Revised Chinchilla scaling laws – LLM compute and token requirements2 people · 2 mentions · first linked 2024-05-01
What Hacker News says?
Paper from a few years ago comparing Google's OCR system to commercially available benchmarks: http://www.educatingsilicon.com/wp-content/uploads/2013/10/p... A lot better for text in photographs. Comparison might be different on dense document text though.
Comment linking /wp-content/uploads/2013/10/photoocr_iccv_paper.pdf
on “Google Cloud Vision API enters Beta”Submitted How much LLM training data is there from all sources in Trillions of Tokens? as a story · 4 points
HN discussionSubmitted Educating Silicon as a story · 1 points
HN discussionSubmitted Revised Chinchilla scaling laws – LLM compute and token requirements as a story · 1 points
HN discussionSubmitted How much LLM training data is there, in the limit? as a story · 1 points
HN discussionSubmitted How much LLM training data is there, in the limit? as a story · 1 points
HN discussionSubmitted Revised Chinchilla scaling implies higher compute requirements for small models as a story · 1 points
HN discussion
