ofermend

10 ms·

91.
by ofermend 3y ago
At Vectara we are exploring what GenAI user interfaces should look like. Vectara-answer is an open source package with our first iteration for question-answering. Would love to hear the community's feedback and would love any contribut
92.
by ofermend 3y ago
I always thought there was something about "I feel lucky" that attracted my attention and now I know what it is: it was the desire to get my answer in a single click, instead of constant searching. That's exactly what's
93.
Vectara Now Integrated into LangChain(blog.langchain.dev)
6 points by ofermend 3y ago|1 comments
94.
by ofermend 3y ago
The Vectara platform combines embedding with a vector store and a SOTA retrieval model, making it a valuable component for building LangChain applications. HN community: what other pain points do you experience when building scalable LLM-p
95.
by ofermend 3y ago
Hey, Ofer here from Vectara. AskNews is a sample application that demonstrates an interface for asking questions about news using the new experience we all expect - ChatGPT like. We first ingest news articles from some selected news sources
96.
Show HN: AskNews – ask questions about recent news(asknews.demo.vectara.com)
5 points by ofermend 3y ago|1 comments
97.
Vectara Releases “Grounded Generation”(vectara.com)
5 points by ofermend 3y ago|0 comments
98.
by ofermend 3y ago
Yes; and the amount of training examples that are really long (relative to the other training examples) becomes small really fast, so it's also a problem of the long-tail in a sense.
99.
by ofermend 3y ago
All the code generation tools, StarCoder included, still have hallucinations. In this context code that looks good, but doesn't work or has a subtle bug. How do we address that?
100.
by ofermend 3y ago
As is typically the case in these situations, it may be easy to build your own (if you have the tech expertise), but then you have to maintain it, upgrade it, and generally keep a team of experts around to make it all work over time. Certa
101.
by ofermend 3y ago
Longer sequence length in transformers is an active area of research (see e.g the great work from the Flash-attention team - https://github.com/HazyResearch/flash-attention ), and I'm sure will improve things drama
102.
Avoiding hallucinations in LLM-powered applications(vectara.com)
145 points by ofermend 3y ago|123 comments
103.
by ofermend 5y ago
Syntegra | Machine Learning Engineer | Onsite/Remote | SF Bay Area | Full-Time | https://www.syntegra.io/ https://www.syntegra.io/jobs/senior-machine-learning-enginee... Syntegra has developed a s