varunkmohan

7 ms·

31.
Why Did Microsoft Build VSCode? Turns Out, GitHub Copilot(codeium.com)
64 points by varunkmohan 3y ago|42 comments
32.
Code Suggestions you Don’t get from Copilot: In-line FIM(codeium.com)
2 points by varunkmohan 3y ago|0 comments
33.
Codeium for Xcode plugin: Free AI autocomplete(github.com)
2 points by varunkmohan 3y ago|0 comments
34.
by varunkmohan 3y ago
Hi from the Codeium team. It's awesome to hear you are allowing other code LLMs to be used on the Replit platform (we're big fans)! We'd love to enable our free chrome extension on Replit.
35.
Samsung bans ChatGPT use due to data leak(bloomberg.com)
2 points by varunkmohan 3y ago|0 comments
36.
Show HN: Codeium-Chrome: Free open-source AI autocomplete Chrome extension(github.com)
2 points by varunkmohan 3y ago|0 comments
37.
Amazon CodeWhisperer is out of beta. We tried it. Spoiler: it isn’t good(codeium.com)
10 points by varunkmohan 3y ago|2 comments
38.
by varunkmohan 3y ago
We explicitly state in our privacy policy that we don't sell user data to third parties. The language server is the common binary cross ides that actually processes local / repo context and communicates with the cloud service to g
39.
by varunkmohan 3y ago
Hi, Varun from the Codeium team here. This is fairly common thing to do to not support these sorts of domains to prevent abuse. Also, Copilot downloads a language server binary as well for their vim extension. Happy to discuss how we could
40.
Show HN: Codeium Chat: Free IDE-Integrated ChatGPT(codeium.com)
3 points by varunkmohan 3y ago|1 comments
41.
Why Your AI Code Completion Tool Needs to Fill in the Middle(codeium.com)
3 points by varunkmohan 4y ago|0 comments
42.
by varunkmohan 4y ago
Sad, this whole thing where companies train chinchilla optimal models has to stop - I'm always happy to get open source models but it would be great if it was useful rather than a benchmark.
43.
by varunkmohan 4y ago
Exactly why our product Codeium (Copilot alternative) supports self-hosting.
44.
AI Should Be Allowed in Interviews. Now It Can(codeium.com)
3 points by varunkmohan 4y ago|1 comments
45.
by varunkmohan 4y ago
OpenAI has done the reasonable thing of not exposing the probability distribution per generated token so it's very hard to use that to completely map to their models. Ultimately, you still do need a very large base model to compete.
46.
by varunkmohan 4y ago
This looks very cool and is a massive improvement. If only OpenAI would also publish what they are doing to make 32K work.
47.
by varunkmohan 4y ago
Curious if anyone has actually used this. It's quite slow for me and feels more like a cute idea rather than a useful product.
48.
Randomness of Generative AI Meets Determinism of Code Syntax Parsing(codeium.com)
3 points by varunkmohan 4y ago|0 comments
49.
Ask HN: Are GitHub clones intermittently slow?
1 points by varunkmohan 4y ago|0 comments
50.
Emacs extension for free Copilot-like AI autocomplete(github.com)
3 points by varunkmohan 4y ago|0 comments
51.
An easier way to search your repo(twitter.com)
1 points by varunkmohan 4y ago|0 comments
52.
by varunkmohan 4y ago
Thanks for the feedback. Currently there is not, but we will look into setting up custom keybindings or some other approach.
53.
by varunkmohan 4y ago
It isn't zero, that's for sure! Before Codeium, we as a team were building scalable ML infrastructure for some of the world's largest ML workloads, so we have a lot of experience in building infra (especially ML serving infra
54.
by varunkmohan 4y ago
I don't really want to argue but we have been pretty open about the fact that we want to keep this product free for users going forward as well. We have been pretty clear about our plan to monetize additional features but the focus now
55.
by varunkmohan 4y ago
We are a free product.
56.
by varunkmohan 4y ago
We are actively iterating on our data cleaning to deanonymize any user information. Of course some cases are tricky to find, but we recognize these issues are serious and will continue to work here. All of our training data is public, so we
57.
by varunkmohan 4y ago
Elixir is not a language we are currently confident in, but in the very near future (O(weeks)), we believe we will have a model that is better than Copilot for Elixir and other niche languages.
58.
by varunkmohan 4y ago
We don't rely on any 3rd party APIs in order to keep serving costs down (which is necessary to keep this as a free service). We only use permissively licensed public code to train.
59.
by varunkmohan 4y ago
No the model actually runs remotely. We would love to have some way to run a lightweight model on the user's machines but it's just not feasible given model size and amount of computation required.
60.
by varunkmohan 4y ago
Sorry to hear that. Point noted, we will take a look - most of our testing has been on chrome / safari.

More ›