tim_sw

17 ms·

241.
How Does Generative Retrieval Scale to Millions of Passages?(arxiv.org)
2 points by tim_sw 3y ago|0 comments
242.
Nvidia shares spike 18% on forecast beat driven by A.I. chip demand(cnbc.com)
38 points by tim_sw 3y ago|14 comments
243.
Anthropic’s Transformer Circuits Publication(transformer-circuits.pub)
2 points by tim_sw 3y ago|0 comments
244.
Interview: Dan Wang, China Specialist(noahpinion.blog)
3 points by tim_sw 3y ago|1 comments
245.
A PreTrainer's Guide to Training Data [pdf](github.com)
2 points by tim_sw 3y ago|0 comments
246.
String-to-String Algorithms from Stanford NLP(github.com)
2 points by tim_sw 3y ago|0 comments
247.
Review: Energy and Civilization(thepsmiths.com)
71 points by tim_sw 3y ago|32 comments
248.
Modernizing Segment's ClickHouse OLAP Platform(segment.com)
4 points by tim_sw 3y ago|0 comments
249.
Meta Custom Silicon: What's Old Is New(semianalysis.com)
6 points by tim_sw 3y ago|1 comments
250.
A New Era for Database Design with TigerBeetle(youtube.com)
4 points by tim_sw 3y ago|0 comments
251.
Meta bets big on AI with custom chips – and a supercomputer(techcrunch.com)
3 points by tim_sw 3y ago|0 comments
252.
GitHub: Working with the LLMs Behind GitHub Copilot(github.blog)
4 points by tim_sw 3y ago|1 comments
253.
Why 2023 is a key year for WebAssembly, a promising cloud technology(runtime.news)
3 points by tim_sw 3y ago|0 comments
254.
SLiC-HF: Sequence Likelihood Calibration with Human Feedback(arxiv.org)
2 points by tim_sw 3y ago|0 comments
255.
by tim_sw 3y ago
We find that responses from existing generative search engines are fluent and appear informative, but frequently contain unsupported statements and inaccurate citations: on average, a mere 51.5% of generated sentences are fully supported by
256.
Evaluating Verifiability in Generative Search Engines(arxiv.org)
1 points by tim_sw 3y ago|2 comments
257.
A List of Leaked System Prompts(blog.matt-rickard.com)
9 points by tim_sw 3y ago|0 comments
258.
Building and Deploying MySQL Raft at Meta(engineering.fb.com)
4 points by tim_sw 3y ago|0 comments
259.
Amazon Plans to Add ChatGPT-Style Search to Its Online Store(bloomberg.com)
2 points by tim_sw 3y ago|0 comments
260.
Language Models of Code Are Few-Shot Commonsense Learners(arxiv.org)
2 points by tim_sw 3y ago|0 comments
261.
Together’s $20M seed funding to build open-source AI and cloud platform(together.xyz)
73 points by tim_sw 3y ago|22 comments
262.
The Acquihire Market for Early Stage Startups Is Ice Cold(hunterwalk.com)
3 points by tim_sw 3y ago|0 comments
263.
OpenAI: Small Teams, Big Impact(sriramkrishnan.substack.com)
2 points by tim_sw 3y ago|0 comments
264.
by tim_sw 3y ago
https://archive.ph/4YOKT
265.
Amazon Overhauls Delivery Network – From National to Regional(wsj.com)
2 points by tim_sw 3y ago|1 comments
266.
First Principles on AI Scaling(dynomight.net)
2 points by tim_sw 3y ago|0 comments
267.
'We Built a Robot That Types': The Man Behind Computerized Stock Trading (2015)(npr.org)
2 points by tim_sw 3y ago|0 comments
268.
The techlash that never happened(eriktorenberg.substack.com)
58 points by tim_sw 3y ago|67 comments
269.
by tim_sw 3y ago
this looks like it's for GCP. TPUs are used for most internal workloads. It's available externally but some of the papercuts and devex without the TPU/TF team helping you can be more painful than using Nvidia/CUDA
270.
GitHub Copilot Leaked Prompt(simonwillison.net)
252 points by tim_sw 3y ago|4 comments

More ›