AI answer engine for complex questions.
San Francisco
Joined June 2022
- Tweets105
- Following23
- Followers8.3K
- Likes125
Phind retweeted
Congrats to @phindsearch on their $10.4M round!
Walls of text are so 2022. On Phind, every answer is now an interactive mini-app.
It’s a step towards their vision of creating a personal internet just for you.
axios.com/pro/enterprise-sof…
We're excited to launch Phind 2 today!
The new Phind is able to go beyond text to present answers visually with inline images, diagrams, cards, and other widgets to make answers much more delightful.
Phind is also now able to seek out information on its own. If it needs more information, it can do multiple rounds of additional searches to get you the most comprehensive answer it can.
Check out our blog post: phind.com/blog/phind-2
Introducing Phind-405B, our new flagship model!
Phind-405B scores 92% on HumanEval, matching Claude 3.5 Sonnet. We're particularly happy with its performance on real-world tasks, particularly when it comes to designing and implementing web apps.
Our focus on technical topics has also made Phind great at answering all questions. We think that our new Phind Instant model, trained similarly to Phind-405B, is the best way to get fast, internet-sourced answers.
More in our blog post: phind.com/blog/introducing-p…
We are excited and proud to be a signatory of SV Angel's Open Letter on AI: openletter.svangel.com/
Phind retweeted
"We find that Phind-70B is in the same quality realm as GPT-4 Turbo for code generation and exceeds it on some tasks.
Phind-70B is significantly faster than GPT-4 Turbo, running at 80+ tokens per second to GPT-4 Turbo's ~20 tokens per second."
What a launch! 👀
Phind retweeted
Introducing Phind-70B – closing the code quality gap with GPT-4 Turbo while running 4x faster phind.com/blog/introducing-p…
Phind retweeted
Phind-70b - A CodeLlama fine tune that claims to have GPT-4 code gen!
Coming soon to open source 👏👏
Phind retweeted
Phind-70B looks like a big deal!
Phind-70B closes the code generation quality gap with GPT-4 Turbo and is 4x faster.
Phind-70B can generate 80+ token/s (GPT-4 is reported to generate ~20 tokens/s). Interesting to see that inference speed is becoming a huge factor in comparing models.
Phind-70B is based on CodeLlama-70B and tuned on 50B tokens more. Supports 32K tokens.
"Phind-70B is also less “lazy” than GPT-4 Turbo and doesn't hesitate to generate detailed code examples."
The thing about custom expert models is that you worry less about hallucinations.
The model weights for Phind-34B will be released soon.
Phind retweeted
I'm very impressed by the way Phind, despite being the tiniest of startups, has managed to keep up with the giants. Phind-70B beats GPT-4 Turbo at code generation, and runs 4x faster. There is definitely still room for startups in this game.
Introducing Phind-70B, our largest and most capable model to date!
We think it offers the best overall user experience for developers amongst state-of-the-art models.
phind.com/blog/introducing-p…
Join us for our San Francisco meetup on February 6th!
We’d love to meet you and hear about how we can keep making Phind better for you.
And, of course, food and drinks will be provided :)
forms.gle/tEGwu2hBtYKptNw98
Announcing much faster Phind Model inference for Pro and Plus users.
Your request will be served by a dedicated cluster powered by NVIDIA H100s for the lowest latency and a generation speed of up to 100 tokens per second.
If you’re not yet a Pro user, join us at phind.com/plans.
🚀 Introducing GPT-4 with 32K context for Phind Pro users.
If you’re not yet a subscriber, join us at phind.com/plans.
Phind retweeted
🆕 pod: Beating GPT-4 with Open Source LLMs
latent.space/p/phind
with @MichaelRoyzen of @phindsearch!
The full story of how Phind finetuned CodeLlama to:
- reach 74.7% HumanEval vs 45% base model
- 2x'ed GPT4 context window to 16k tokens
- 5x faster than GPT-4 (100 tok/s)
all with just 70B extra tokens and a tiny team, shooting right to the top of the HF leaderboard!
plus: what it's like to have @paulg and @RonConway as investors!
Phind retweeted
Phind - @phindsearch worked pretty well for my tiny Bash script. It even adopted how I use a third-party script (jq).
All on first try.
See it in action:
Phind can now beat GPT-4 at programming, and does it 5x faster.
phind.com/blog/phind-model-b…