fzliu

27 ms·

511.
UUIDv7(rfc-editor.org)
2 points by fzliu 2y ago|1 comments
512.
by fzliu 2y ago
Agreed. ViTs are better if you're looking to go multimodal or use attention-specific mechanisms such as cross-attention. If not, there's evidence out there that ViTs are not better than convnets for small networks and at scale ( h
513.
Painting the Town AI(sequoiacap.com)
1 points by fzliu 2y ago|0 comments
514.
by fzliu 2y ago
A lot of folks I've spoken with say that single-agent systems are still extremely limited, let alone multi-agent platforms. In general, it seems to boil down to: - Agents need lots of manual tuning and guardrails to make them useful -
515.
Estimating the Doubling Rate(misha.brukman.net)
2 points by fzliu 2y ago|0 comments
516.
Community Is a Tamagotchi(remotesynthesis.com)
4 points by fzliu 2y ago|0 comments
517.
LlamaIndex on Vertex AI(cloud.google.com)
2 points by fzliu 2y ago|0 comments
518.
Big Companies vs. Startups (2015)(danluu.com)
49 points by fzliu 2y ago|59 comments
519.
Cohere Int8 and Binary Embeddings(cohere.com)
1 points by fzliu 2y ago|0 comments
520.
Did GitHub Copilot increase my productivity?(trace.yshui.dev)
244 points by fzliu 2y ago|320 comments
521.
Drug Discovery with Vector Search(github.com)
17 points by fzliu 2y ago|1 comments
522.
by fzliu 2y ago
> I think a few details (and perhaps a small amount of customization) would go a long way. I hear you and agree 100% - I unfortunately haven't gotten around to writing better documentation nor solid code samples that utilize Radient
523.
Radient – vectorize many data types, not just text(github.com)
63 points by fzliu 2y ago|14 comments
524.
What Is the Raft Consensus Algorithm?(yugabyte.com)
2 points by fzliu 2y ago|0 comments
525.
Developer Company Blog Posts(plushcap.com)
1 points by fzliu 2y ago|0 comments
526.
Using Your Vector Database as a JSON (Or Relational) Datastore(zilliz.com)
43 points by fzliu 3y ago|20 comments
527.
Runhouse(github.com)
1 points by fzliu 3y ago|0 comments
528.
Vectorizer(github.com)
2 points by fzliu 3y ago|0 comments
529.
Tokens, n-grams, and bag-of-words models (2023)(zilliz.com)
160 points by fzliu 3y ago|16 comments
530.
by fzliu 3y ago
> You're double-dipping into the data. You look at the performance, then you tune some part of the network by hand, see if it helps, and then keep doing that. It's testing on the training data. I purposely tried to avoid adding
531.
by fzliu 3y ago
I recommend checking out this paper: https://arxiv.org/pdf/2310.16764.pdf . From the concluding section: "Although the success of ViTs in computer vision is extremely impressive, in our view there is no strong evid
532.
Vision Transformers Are Overrated(frankzliu.com)
67 points by fzliu 3y ago|17 comments
533.
Machine Learning, AI and Data Landscape(mattturck.com)
2 points by fzliu 3y ago|0 comments
534.
Tokens, N-Grams, and Bag-of-Words Models(zilliz.com)
2 points by fzliu 3y ago|0 comments
535.
x-transformers(github.com)
1 points by fzliu 3y ago|0 comments
536.
Sequence-to-Sequence Toolkit Written in Python(github.com)
1 points by fzliu 3y ago|0 comments
537.
Vision Transformers Are Overrated(frankzliu.com)
2 points by fzliu 3y ago|0 comments
538.
Powerlifting and Brexit (2016)(alexgraveley.com)
3 points by fzliu 3y ago|1 comments
539.
Vision Transformers are Overrated (2023)(frankzliu.com)
2 points by fzliu 3y ago|0 comments
540.
Scale Was All We Needed, at First(aiacumen.substack.com)
1 points by fzliu 3y ago|0 comments

More ›