varunkmohan

13 ms·

91.
We built a giant LED sign on Castro St(twitter.com)
2 points by varunkmohan 4y ago|0 comments
92.
by varunkmohan 4y ago
Yes, that's on the list of plugins to create. We're keeping folks updated on the timelines on our Discord.
93.
by varunkmohan 4y ago
We are currently working on making suggestions more tailored to your codebase without having to persist / upload the codebase. We will have updates on this shortly!
94.
by varunkmohan 4y ago
Interesting - would love if you could join our Discord ( https://discord.gg/3XFf78nAx5 ) and we are happy to help figure out what's going on!
95.
by varunkmohan 4y ago
Thanks for the feedback!
96.
by varunkmohan 4y ago
Please join our Discord ( https://discord.gg/3XFf78nAx5 ) and we are happy to help debug this for you!
97.
by varunkmohan 4y ago
Thanks for the support! We are actively looking to hire experts in program synthesis and ML infrastructure to help deliver this technology at scale
98.
by varunkmohan 4y ago
We have gotten requests for more IDEs, and are actively working on creating support for them so that as many developers as possible can leverage this technology!
99.
by varunkmohan 4y ago
We're actively thinking of ways where we host the service within a customer's tenant - more coming soon!
100.
by varunkmohan 4y ago
We understand the concern! We describe our security policies in detail on our website, but we offer users the option to opt out from code snippet telemetry. This means after the inference, we do not store any code-based information on our e
101.
by varunkmohan 4y ago
Thanks! We've made a lot of optimizations internally for latency. We've gotten a lot of feedback that it is faster than Copilot but obviously they operate at much larger scale.
102.
by varunkmohan 4y ago
Right now, we are providing a free alternative to Copilot and expanding the set of features we support. We are actively working on improving the quality of suggestions and would appreciate feedback. We guarantee that all users that join now
103.
Show HN: Codeium – a free, fast AI codegen extension(codeium.com)
86 points by varunkmohan 4y ago|39 comments
104.
Loading 60 BERTs onto a single T4(exafunction.com)
3 points by varunkmohan 4y ago|0 comments
105.
Getting Rid of CPU-GPU Copies in TensorFlow(exafunction.com)
6 points by varunkmohan 4y ago|0 comments
106.
by varunkmohan 4y ago
Completely agreed. For some of these large language models, it would take a long time before inference spend dominates training spend.
107.
by varunkmohan 4y ago
I think you'd probably always want to go with T4's since they are the same price unless there's just no availability for them.
108.
by varunkmohan 4y ago
Yeah, we should have a public release very soon for people to deploy internally. We will have support for all the commonly-used frameworks and different versions.
109.
by varunkmohan 4y ago
Yup, exactly. It's a good point that for self-supervised workloads, the training set can become arbitrarily large. For a lot of other workloads in the vision space, most data needs to be labeled to be able to used for training.
110.
by varunkmohan 4y ago
rCUDA is super cool! One of the issues though is for a lot of the common model frameworks are not supported and a new release has not come out a while.
111.
by varunkmohan 4y ago
You have a good point. I think for small enough workloads self managing instances on-prem is more cost-effective. There is a simplicity gain in being able to scale up and scale down instances in the cloud but may not make sense if you can s
112.
by varunkmohan 4y ago
Disclaimer: I'm the Cofounder / CEO at Exafunction That's a great point. We'll be addressing this in an upcoming post as well. We've served workloads that run entirely on spot GPUs where it makes sense since a small
113.
by varunkmohan 4y ago
Agreed that there are workloads where inference is not expensive, but it's really workload dependent. For applications that run inference over large amounts of data in the computer vision space, inference ends up being a dominant porti
114.
Are GPUs Worth It for ML?(exafunction.com)
131 points by varunkmohan 4y ago|91 comments
115.
by varunkmohan 4y ago
Hi everyone! My name is Varun Mohan and I'm the CEO of Exafunction (we're hiring!). We're excited to share a bit about our platform, which we think will be the best way to run deep learning and GPU workloads in the cloud, sta
116.
Show HN: Exafunction, efficient deep learning at scale(exafunction.com)
8 points by varunkmohan 4y ago|5 comments
117.
by varunkmohan 5y ago
For the public API, we should be under 100ms per token but don't have a strict guarantee. If you have a strict SLA, you can talk to us and we can get it to as low as 20ms per token at high load.
118.
by varunkmohan 5y ago
Unlike OpenAI, which is serving a closed-source model, anyone can validate for themselves (even without using our service) whether open-source models like GPT-J are suitable for a particular application. We'd like to work with the open
119.
by varunkmohan 5y ago
Great suggestion! We will look into it.
120.
by varunkmohan 5y ago
There's some secret sauce behind it, but mostly just using relatively inexpensive cloud inference hardware very effectively. It turns out most of the common NLP frameworks leave a good deal of performance on the table, not to mention t

More ›