Unify
@letsunifyai
Followers
3,012
Following
76
Media
121
Statuses
251
The Best LLM on Every Prompt ✨
London
Joined January 2021
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
#LingOrmFanMeet2geTHer
• 277631 Tweets
DEAR MYLOVE LINGORM
• 272558 Tweets
Edmundo
• 126511 Tweets
#ZeeNunew1stFMinTokyo
• 87164 Tweets
#KPOPMASTERZinBANGKOK2024
• 41914 Tweets
Oregon
• 31426 Tweets
#sbhawks
• 27184 Tweets
京成杯AH
• 25001 Tweets
セントウルS
• 22428 Tweets
アスコリピチェーノ
• 12737 Tweets
メンデス
• 10445 Tweets
Pinned Tweet
Very excited to release our Model Hub 😍 - a collection of LLM endpoints with live runtime benchmarks all plotted across time 📈
We currently have 21 models provided by:
@anyscalecompute
,
@perplexity_ai
,
@replicate
,
@togethercompute
,
@octoml
,
@MistralAI
15
70
318
We are excited to release Ivy, a new open source Deep Learning framework!
Ivy unifies the syntax and call signatures of existing frameworks.
Write your code once in Ivy, and support all frameworks simultaneously.
Links to Paper, Code and Docs here:
26
76
334
Happy to announce Ivy Robot!
The library offers a variety of functions for gradient-based motion planning.
Use with mobile robots or manipulators!
Every function is jointly compatible with TensorFlow, PyTorch, JAX, MXNet, and NumPy.
Docs, code + demos:
0
12
82
📢 We're thrilled to announce that Ivy has been selected for
@GoogleOSS
Summer of Code 2023! 🎉 👏
We'll be mentoring students to work on unifying ML by developing backend for various ML libraries and implementing DL models in Ivy 🧠 💻
More info and links below 🧵
⬇️⬇️⬇️⬇️⬇️
2
8
50
Happy to announce Ivy Gym!
The library offers fully differentiable classic control tasks 🤖
Bypass value functions, and minimize the reward directly ✅
Each task is jointly compatible with TensorFlow, PyTorch, JAX, MXNet, and NumPy
Docs, code + demos:
0
2
45
We're on a mission to unify all ML frameworks, making them all seamlessly work together, without any code changes 💫
Fancy joining us on this ambitious journey? 🧑💻
We're hiring all over the world! 🌎
Hear what our wonderful team have to say 😊
3
11
44
Huuuge honour to be accepted into the first ever
@IntelIgnite
London batch 😍 🎉
@intel
shares our vision for unified AI 🟢, with great initiatives like OpenVINO and OneAPI. We're incredibly excited to now be working with some of the engineers behind these great tools 🧑💻, as we
4
15
43
We've just passed 1000 contributors 📈💪and we're one of the world's top OS companies according to OSSRank 🧑💻😍
Huge congratulations to all 1000 contributors who have made this happen 👏
Stay tuned for our public release, only a few weeks away! 👀
1
18
44
Introduing Ivy Modules!
Define your models and optimizers in Ivy, and then train using any of: PyTorch, TensorFlow, Jax or MXNet!
Ivy does not wrap classes from these frameworks.
Ivy classes build directly on Ivy's simple functional API.
Find more at:
0
9
38
We had so much fun at
@theaiconference
last week! 😍 Special thanks to
@hwchase17
(CEO
@LangChainAI
),
@nazneenrajani
(Research Lead
@huggingface
), Andy Hock (VP Product
@CerebrasSystems
),
@pirroh
(VP of AI
@Replit
) and
@Azade_na
(Research Scientist
@GoogleDeepMind
) for taking the
0
6
37
🎉 Very excited to sponsor this year's
@AIConference
!
🌟 Expect powerhouse talks by AI pioneers from
@NVIDIAAI
,
@Google
,
@huggingface
, and more!
Stay tuned for a sneak peek into our poster! 🤩
#AITalks
#aiconf2023
1
5
36
We're happy to announce that End-to-End Egospheric Spatial Memory (ESM) will be presented at
#ICLR2021
this Thursday, 9-11am PST 🥳
ESM was implemented using Ivy, supporting all modern DL frameworks ✅
Easily drag-and-drop ESM into your own code 🙃
Link:
I'm very excited to be presenting our work: End-to-End Egospheric Spatial Memory (ESM) at
#ICLR2021
this Thursday, 9-11am PST!
ESM encodes the memory in an egosphere around the agent, enabling expressive, efficient and structured local 3D representations:
4
24
148
0
6
34
Very Excited to announce that we have been nominated for hottest AI startup 2023🔥🤖 in the Europas Awards 🌍🎉
With so many great nominees this year, we are honoured to be considered amongst them⚡
Voting is open and every vote for Ivy would be greatly appreciated, you can
2
7
36
Super excited to be chatting with
@HowDevelop
about our
#Ivy
#GSoC
projects this Friday (10th March), hosted by
#scaler
! 🎉 😊 All are welcome! 🧑💻
We'll be diving deep 🤿 into each of our ML projects 📉, more info at:
Hopefully see some of you there! 😍
0
10
33
We've just secured £325k in grant funding from the UK Government 🎉😍
Huge honour for
@innovateuk
to support our ongoing research 🧑💻
The project will enable source code generation in the target framework when converting 🔁🗒️
Let's keep unifying! 💪🟢 😁
2
5
32
Huge honour to present alongside GitHub Cofounder
@chacon
, Chief Decision Scientist at Google
@quaesita
, MongoDB + Business Insider founder
@kevinryan
, and many other amazing speakers. Check out our Keynote at J12 Day! 📽
We explain the fragmented ML
2
9
33
Very excited to announce our brand new Reading Group YouTube Series, which kicks off with a livestream tomorrow at: 09:30 PT, 12:30 ET, 16:30 GMT, 18:30 CET, 22:00 IST [] !👨💻
Dan FU [] will be presenting his great work on Flash
0
9
34
Excited to announce our new contributor program! 🎉☺️ Contributors will get the chance to work alongside our team, get 1-on-1 mentorship, and join in with our meetings and discussions. We’ve also created a series of badges🛡️ for different types of contributions 🧑💻, which will be
0
8
30
Introducing Implicit Rendering with Ivy Vision!
Drag-and-drop key functions into your own projects✅
As a demo, we implement Tiny-NeRF with simultaneous support for TensorFlow, PyTorch, JAX and MXNet.
More functions (and high-res demos) coming soon!
0
4
33
Such a pleasure joining Andrew on
@churnfm
📻, thanks for having us! ♥️ We talk community 🫂, Generate AI 🤖, open source 🧑💻, and much more! Stay tuned for more updates coming soon 📡👀 Happy Wednesday! 😁
Today on the show we
@DanielLenton1
, CEO and Co-Founder of
@letsunifyai
Listen to Dan's unusually powerful way that they built their community at Ivy:
0
0
2
0
10
31
Hello world. Time to break a year of radio silence😁
We're working hard on release 1.2, and our first transpiler demo!🔁
In the meantime, feel free to hop on any of the weekly "Ask us Anything!" meetings with our engineering team🔧
All are welcome🙂
2
11
29
In 20 minutes we will be taking part in a live open discussion "The Next Frontier: Where Does AI Go, How Do We Shape an Innovative Future?" 🤖. You can sign up for this free event right now [], everyone is welcome, regardless of your prior experience! As
0
7
29
We've opted-in for
#Hacktoberfest2022
!
Whether you’re a veteran🥷or a beginner🤷:
· Register:
· Join:
· Read:
· pick an issue with the ‘hacktoberfest’ label
+Let’s get Hacking! 🧑💻
2
8
22
It was an absolute joy to chat with
@emilsnotes
last week, as part of their awesome new
@kapa_ai
podcast 🎙 😊
We talked about open-source 📜, community building 🧑💻, and the journey for first-time founders 🗺
Go check out their podcast, and check out our video if you'd like to
how do you got from idea to +9k GitHub stars and a thriving dev community in <2 years? 👀
here's a clip from the newest
@kapa_ai
podcast with
@DanielLenton1
from
@letsunifyai
who is on a super ambitious mission to unify all ML frameworks 💥
1
0
4
2
9
29
Join the RoundTable Chatroom at NeurIPS 21 next Friday 8pm UTC!🗣
We have an amazing list of speakers and topcis🤖⚡
Places going fast!💥Sign up now to save your spot!⬇️
Ivy are proud sponsors of open and inclusive discussion on important ML topics💬🤔
🥳 We are proud to present our second event of
#RoundtableChatroom
at
#NeurIPS2021
@NeurIPSConf
sponsored by
@ivythread
!
Save the date: 8-10PM GMT FRI 10 DEC, 2021 (ZOOM)
Sign-up:
Each discussion topic will be guided by a knowledgeable Mentor below ⬇️
1
2
10
1
7
28
@fchollet
Great to see Keras returning to it’s unifying roots! 💪😍 With this new avenue to run pure-Keras models in PyTorch + JAX, and with Ivy providing tools to convert in all other directions as well, AI developers now have a great suite of tools at their disposal for cross-framework
0
7
24
Need 3D Vision functions in your end-to-end graph?
Ivy Vision offers differentiable functions for rendering, voxel grids, SDFs, optical flow, depth triangulation, cost volumes and more!
Check out the docs and demos!
Docs:
Code:
We are excited to release Ivy, a new open source Deep Learning framework!
Ivy unifies the syntax and call signatures of existing frameworks.
Write your code once in Ivy, and support all frameworks simultaneously.
Links to Paper, Code and Docs here:
26
76
334
0
5
25
If you'd like to learn more about our roadmap at Ivy 🟢, everyone is welcome to join the (free) PyData event starting in 2.5 hours! (19:00 CET) ✨
We love answering questions at all levels, so if you have any, then please join in and ask away 😊
0
9
22
Happy to announce Ivy Mechanics!
The library offers differentiable functions for orientation + pose conversions, co-ord transformations, and more!
Every function is jointly compatible with TensorFlow, PyTorch, JAX, MXNet, and NumPy.
Docs, code + demos:
0
1
25
Can't wait! 😊
During our 90-minute workshop at ODSC West, we will show how Ivy can instantly transpile ML models between different ML frameworks, using SOTA transformer models as examples 🔁
If this sounds interesting, then please save the date! November 1st - 3rd🗓️
#ODSCWest
Instructor Spotlight - Daniel Lenton. Join at West and learn about
IVY - The unified machine learning framework, and its mission to unify all machine learning frameworks.
@letsunifyai
Register now:
0
0
0
1
7
19
Super excited to be presenting at ODSC East later today 😍
We'll be sharing early previews to some live demos, ahead of our upcoming public pilot release 🚀
If you're attending, come join us at 12:20 EST:
Stay tuned for more updates 📺 - happy Tuesday!
0
8
23
Huge honour to join
@sramji
on his Open||Source||Data podcast 🎙️, which is out now! (link below ⬇️). We talk about the future of AI 🤖, open source 🧑💻, and even a little on guitars 🎸. For someone with such an impressive track record, Sam is incredibly down to earth, and chatting
A new episode is here!
@sramji
and
@DanielLenton1
discuss the power behind Ivy , open source reproducibility, and democratizing AI.
Tune in:
#OpenSourceData
#Ivy
#Reproducibility
#DemocratizingAI
0
2
3
0
4
24
SORA from OpenAI is amazing🤯 We just posted a video explaining how we "think" it might work, and how it compares to the very recent Lumiere from Google. Feel free to check it out 😊 Exciting time for T2V! 🔥
0
5
24
Thanks Oliver! ☺️ Keep up the great OS work at Red Hat ⛑💪
If you're into
#ML
- you should take a look at these folks and follow them. Thanks
@DanielLenton1
and
@ronnieclark__
for pointing me to it!
#MachineLearning
0
0
3
4
0
18
We're very excited to announce that Unify is co-organising "The AI Showcase", in partnership with
@AWSstartups
,
@TYTN_ai
, and
@OctopusVentures
, in London on 28th November 🚀😍
Connect with London’s bustling AI community, learn from industry experts, and see live demos from the
0
4
22
Very excited to welcome
@sanchitgandhi99
, Audio Team Lead🎙️at
@huggingface
🤗, to our livestreamed reading group in 3 hours!
He will be presenting Distill Whisper, where they used Knowledge Distillation to compress Whisper, making it ~6 times faster 🤯
He also developed
0
4
22
Thanks for sharing! 😊
TIL about a very interesting project by
@letsunifyai
() that attempts to transliterate seamlessly between different
#AI
/
#MachineLearning
model implementations.
Go take a look at what they are proposing, and their repo at:
0
2
4
0
0
16
🔹🔷
@anyscalecompute
is consistently among the fastest providers for both llama-2-70b-chat and mixtral-8x7b-instruct-v0.1 in tokens per second ⚡, and they have the best handling for concurrent requests on llama-2-70b-chat. They are also similar to Together AI tokens per second
1
2
22
Very excited for this "Lightning Interview ⚡️" with ODSC founder Sheamus McGovern in a few days!
Tune in if you'd like to hear more about what we're up to at Ivy 🤔🔧🟢🧑💻🦾
All are welcome to listen in and ask questions 😊
No question is too small!
On August 12th, we will interview ODSC West speaker Daniel Lenton – CEO at Ivy. He will lead an engaging conversation on "Ivy - The Unified Machine Learning Framework". Listen in and ask questions, live!
@letsunifyai
0
2
4
0
8
19
Huge honour to present Ivy at
@NeurIPSConf
last week! Thanks to everyone who came along and asked so many great questions 🙏❤️. Some of our top picks at the conference were:
Are Emergent Abilities of Large Language Models a Mirage? [] 🔥
Scaling
1
7
20
Had a lot of fun chatting with Digital First last week! Thanks for having us 🙏😊 Full article below ⬇️ We discuss where AI in heading 🤖, how companies should be leveraging it 📈, and the challenges of being a founder 🧑💻. More announcements are around the corner, so stay tuned
0
4
20
We had a blast at the
#GSoC
Mentor Summit last weekend! 😍🎉 Such an honour to meet so many open source rockstars🧑💻🎸, the famous global "Chocolate Table" was definitely a highlight 🍫. Can you guess what country Ivy is representing? 🌎👀 Pretty difficult right? 😅
1
3
21
Want to compare outputs across all the latest LLMs, side by side? Check out our shiny new (and totally free) chat interface ✨ Tell us what you think!
Happy Prompting 🤖
0
10
20
We're honored to be the first interview for the new OSS Spotlight Series by Decibel Partners 😊
We had a great chat with Sudip, and they've done a great job writing this up.
To learn more about how we're trying to unify all ML frameworks 🟢, feel free to go and check it out! 👀
Excited to kick off the
@DecibelVC
OSS Spotlight Series where we interview creators of really interesting open-source projects! In the first installment, we spoke to
@DanielLenton1
about his vision to unify Machine Learning frameworks:
0
4
10
1
3
15
On top of Ivy, we release four Ivy libraries:
Mechanics
3D Vision
Robotics
Gym Environments
Each function can be "dragged-and-dropped" straight into any project written in: PyTorch, TensorFlow, JAX, MXnet, or Numpy.
Try them out, and let us know what you think!
1
0
17
We're honoured to be listed by
@RunaCapital
as one of the fastest-growing OS projects in Q3 2022! 📈 ☺️
Huge shout out 📢 to our devoted community of 540 contributors 🧑💻 and 7.4k star gazers ⭐️, who have all made this milestone happen! 💪
We're just getting started! 🟢⚡️😁
Top 10 fastest-growing open-source projects in Q3 2022 👇
1.
@builderio
2.
@novuhq
3.
@letsunifyai
4.
@turbothq
5.
@apify
6.
@qovery_
7.
@deno_land
8.
@flipper_zero
9.
@penpotapp
10.
@rustdesk
5
20
62
2
6
18
Hello World 🤖... We're Hiring! 🧑💻
We're looking for a Full-Stack Staff Engineer to join our mission to unify all ML frameworks 🟢✨
You will lead the charge for our SaaS offering and UI, and manage a growing team spread across the globe 🌎
More at:
1
11
19
New Feature ✨ You can now specify keywords like "highest-tks-per-sec" or "lowest-input-cost" in our API, and we'll automatically route to the best provider. Try it out and let us know what you think!
Happy Prompting 🤖
0
7
19
🤖 Build your own Application. In the video below we show how you can get started exploring the dashboards, and using single sign on to access all models. You can easily switch models midday through a conversation, to optimize to better speed ⚡, cost 💸, or both 🔥
1
1
19
All of this data is constantly evolving, and trends that hold true right now might not hold true next month 🗓️, next week ⌛ or even the next day ⌚, either due to changing traffic, networking speeds, API or hardware updates behind the scenes, or many other factors 🤷. For the
0
0
18
Thanks for the contribution and for the shout-out Kanak! ☺️
All applications to our open roles include a short paid open-source coding challenge, which we guide you through 🧑💻
If you’re interested in joining our team, then all are welcome to apply! 🟢
I have been contributing to open source projects for a long time, but have never been paid before. Well, that changes today.
Thank to
@DanielLenton1
@letsunifyai
for the opportunity. I hope I can join your team some day.
#OpenSource
#github
#techtwitter
#MachineLearning
1
0
4
2
1
12
Really excited to announce
@luka_ribar
and
@savelichic
from
@graphcoreai
as our guests for today's live reading group! 😊They will be presenting "SparQ Attention: Bandwidth-Efficient LLM Inference", which is a technique for increasing the inference throughput of LLMs by reducing
At 2pm today Graphcore researchers
@luka_ribar
&
@savelichic
will be presenting at
@letsunifyai
's popular reading group.
We'll be covering our recent SparQ paper - a method for increasing LLM inference throughput by sparsifying attention.
Live stream:
0
4
7
0
3
17
Glad our sponsorship helped! 🎉 Honoured to be a part of this great event ☺️
Hosting GDSC WOW Mumbai with 250+ people was a huge success! 🎉
Thank you to everyone who made this weekend's event unforgettable, especially
@letsunifyai
and
@letsupgrade
@GdscWowMumbai
@gdscwow
#events
#community
3
5
85
0
3
17
Excited to dive into "The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits" in today's reading group 😁. It matches the performance of full precision transformers, despite being much cheaper, faster, and using less memory 🔥
If you'd like to learn more, then we'll
1
4
17
You've already been representing Ivy out there and were already a part of the team 😉, but welcome Suvaditya! (and thanks for the shout-out!) looking forward to unifying AI 🚀
I'm officially joining
@letsunifyai
as an ML Research Engineer Intern today! (Gonna work on some really cool projects here)
So excited to unify all ML frameworks together into one!
Go take a look at now and join me on my journey. See where it takes you!
15
2
43
0
1
17
Excited to welcome
@fe1ixxu
to this weeks reading group! 📽️ 😊 He will be presenting his recent work on Contrastive Preference Optimization (CPO), beating GPT-4 on various WMT datasets 🔥👀
You can find the link for the livestream here []
As always, we
3
2
15
Excited to welcome
@AshkboosSaleh
to this week's reading group ☺️. He will be presenting SliceGPT 🤖, which can remove up to 25% of the model parameters for LLAMA-2 70B, OPT 66B and Phi-2 models while maintaining 99%, 99% and 90% zero-shot task performance of the dense model 🔥👀
0
6
16
Some LLMs are very sensitive to simple prompt tweaks. See below how Gemma 7b performs when asking it to find the "largest/highest/maximum" number from three inputs.
When prompted with "largest", it incorrectly sorts the numbers as strings, while the other prompts yield the
0
4
16
Enjoyed having your there
@acecornball
! Thanks for the mention.
For anyone else interested in learning a bit more about what we're up to, we are having these 30-60 minute meetups every weekend, all are welcome!
Check out our discord if interested 🙂
#100DaysOfCode
Day : 52/100
coded for 3 hours in sessions. prepared for exam
attended a online community meetup
@letsunifyai
#mlh
#github
#coding
#opensource
#unifyai
#ivy
0
2
5
0
2
15
🟰
@replicate
presents another well-balanced option for llama-2-70b-chat and mixtral-8x7b-instruct-v0.1 for input-heavy tasks 🤝. They are more cost effective than other providers such as Anyscale or Together AI when it comes to input prices, while being faster than other
1
0
15
📈 Plotting across time is essential. Considering the example for llama-2-70b-chat below, if we were to measure the output tokens per second once at 2am on Feb 3rd, you would find Together AI were the slowest provider. However, if you took the exact same measurements only 12
1
1
16
Can't wait to livestream 🎥 our webinar on "Democratizing AI Deployment" at Data Phoenix! 😍
We will dive into the complexities of AI deployment 🛠 , diving into orchestration 🔀, compression ⬜ ➡▫, compiler infrastructure ⏬, and hardware ⚡. We will also show how to reach
0
3
15
Excited to welcome Chengming Hu to our next reading group session! 🌟We will be diving into his recent survey on Teacher-Student Architectures for Knowledge Distillation 🤖⬅️📚
Join us live today (at 7:00 am PT, 10:00 am ET, 3:00 pm GMT, 4:00 pm CET, 8:30 pm IST).
Whether
0
3
15
Excited to announce that we will welcome
@ZhengxuanZenWu
and
@aryaman2020
from
@stanfordnlp
to our Paper Reading Group on 24th May! ✨
They will be presenting their paper, "ReFT: Representation Finetuning for Language Models" 🔥
Sign up 👉
Instead of
1
7
15
🇲
@MistralAI
has the fastest TTFT for the mixtral-8x7b-instruct-v0.1 model in many regimes ⚡, making it a great option for those looking to build highly responsive applications on top of their model. The input cost is also similar to other offerings, making it a cost effective
1
0
15
Thanks for the kind words Mayank 🙂, and thanks for the important contribution! 🎉💪
Hey folks, I just Contributed in Revolutionary idea that is gone make development more efficient in ML, Got my PR merged😃 at
@letsunifyai
created
@TensorFlow
fronted function
Thanks
@kunalstwt
for introducing
#OpenSource
#community
#code
#github
#MachineLearning
#Unity
4
3
79
2
0
10
Really excited to announce that Suvaditya will be presenting some of our latest work at
#DevfestNewDelhi
tomorrow! 💥
See the original post for more details 🟢
Hopefully see some of you there! 🧑💻😁
We're excited to announce that Suvaditya Mukherjee, an ML research engineer intern at IVY, will be joining us for Devfest New Delhi tomorrow! He will be informing us about what is IVY and how it amplifies all the ML frameworks together.
2
4
18
0
5
13
We had sooo much fun at ModCon last week 😍
Thanks to
@clattner_llvm
,
@iamtimdavis
and
@jeremyphoward
for explaining their views of the AI landscape (we shamelessly got some selfies 📸 😁)
With so many amazing new AI companies and tools coming onto the scene, we really are
0
4
13
We had so much fun on "The Accelerometer" podcast 🎥🎙️😍, thanks for having us 🙏 We talk about open source 📖, the impact of AI 🤖, and how to hire great talent 🧑💻. Link to the full video below ⬇️. Hope everyone is having a wonderful festive period ✨, and a great end to the
0
3
13
Thanks
@aman_shakesbeer
! 🙏
This will definitely make the lives of all our M1 users a lot simpler 😅
Made my first contribution to
@letsunifyai
. Now, M1-based MacBooks can convert ivy codes to Mxnet without Docker.
I genuinely believe that the polymorphic ability of Ivy on a powerful chip like M1 will indeed be a game-changer for ML Engineers and researchers.
0
0
5
0
1
7
⚙️
@perplexity_ai
represents a middle of the market option for llama-2-70b-chat 🤝, especially when it comes to input-heavy tasks such as document summarization 🗂️. They are faster than OctoAI, whilst being more cost effective than either Anyscale or Together AI, striking a good
1
0
14
🐙
@octoml
has the lowest input cost for both llama-2-70b-chat and mixtral-8x7b-instruct-v0.1, making it the most cost effective choice for input heavy tasks such as document summarization 🗂️. It also has the lowest output cost for mixtral-8x7b-instruct-v0.1, making it also a
1
0
13
Ivy was used to implement End-to-End Egospheric Spatial Memory (ESM), accepted as a conference paper at ICLR 2021.
ESM encodes the scene into an egosphere, which travels with an agent.
The official Ivy implementation will be released soon!
0
0
11
Excited to be livestreaming 🎥 at Data Hour later today (13:30 GMT, 14:30 CET, 19:00 IST), as we talk about our vision to Democratize AI Deployment 🤖🟢⚡️
Feel free to join us! 🧑💻 It's totally free, and we love answering questions - both big and small ☺️ See you there! ✌️❤️
0
2
13
The low level Ivy API keeps the developer in full control of their code, allowing complexity to be abstracted only where useful.
Ivy wraps the native framework, which itself wraps the efficient C++ backend. All these levels of abstraction remain accessible to the developer.
1
0
11
Really excited to present in front of the
@PaddlePaddle
community tomorrow😊
We will show how Ivy can convert any
@PyTorch
model from
@huggingface
to Paddle Paddle, in one line of code ✨
The meeting is open to everyone, so feel free to attend if you'd like to learn more!
0
3
13
Webinar 01 of our Weekly Webinar Series is up on the Unify YouTube channel in case you missed it! ✨
This week
@DanielLenton1
showcased how you can build a chatbot using the Unify API and went in-depth on when to use synchronous and asynchronous clients 🤖
Check it out -
0
5
13
LLM Contributor Projects Week - LlamaIndex RAG Playground🛝
Abhijeet Chintakunta has developed the LLamaIndex RAG Playground, an application that allows users to upload PDF documents and ask questions about them!⚡️
Try it out -
Abhijeet used LLamaIndex
0
2
12
🫂
@togethercompute
has the fastest implementation of mixtral-8x7b-instruct-v0.1 among all the providers tested, regularly reaching as high as 90 tokens per second ⚡. They also have the lowest output cost for llama-2-70b-chat, making it a great choice for output heavy tasks such
1
0
12
We're really excited for this one! See you there 👉
Let's talk about the shortcomings of traditional naive RAG systems and explore how advanced agentic RAG systems effectively address and overcome these challenges.
Join
@atitaarora
and
@iamjameskeane
for our joint Webinar with
@letsunifyai
.
RSVP: 🚀
0
5
15
0
3
12
Here we show how various Ivy functions can be combined for gradient-based motion planning of a drone in a cluttered scene.
This is just one example, check out the demos to see what else is possible with the Ivy libraries!
1
0
10
Hands down most impressive demo this year, huge congrats to the
@cognition_labs
team 🔥 "Becnhmark the performance of Llama on a couple of different API providers" sounded familiar 👀
I wonder whether Devin knows about the various live Llama benchmarks available?
Today we're excited to introduce Devin, the first AI software engineer.
Devin is the new state-of-the-art on the SWE-Bench coding benchmark, has successfully passed practical engineering interviews from leading AI companies, and has even completed real jobs on Upwork.
Devin is
5K
11K
45K
0
1
12
You read our minds. This miiiight be on our roadmap...
I'd love to see this, but for stable diffusion providers. I'm pretty sure
@fal_ai_data
is the best in terms of both performance, price and E2E latency but happy to be proven wrong and get more motivation to optimize it even further!
1
0
12
0
3
12
Happy to have you on the team! 😊 and looking forward to unifying AI together! 🚀🚀🚀
0
2
11
Happy Friday ✌️
Don't forget you can compare all the new models released this week on our new chat interface for free!
Try it out here👉
0
4
10
Thrilled to announce
@XueFz
as our guest for today's live reading group! 😊 He will be presenting "OpenMoE: An Early Effort on Open Mixture-of-Experts Language Models". They show how cost effective MoEs can be 💸, and provide in-depth analysis of the ideal routing mechanism 🔀
(1/5)🚀 Our OpenMoE Paper is out! 📄 Including:
🔍ALL Checkpoints
📊 In-depth MoE routing analysis
🤯Learning from mistakes & solutions
Three important findings:
(1) Context-Independent Specialization;
(2) Early Routing Learning;
(3) Drop-towards-the-End.
Paper Link:
5
105
516
1
5
11
We are really excited to announce that we will be joined by
@CopilotKit
for our Webinar Series this Thursday!🤩
RSVP 👉
In this session, we'll explore how CopilotKit makes integrating AI-powered Copilots into any product easy. We'll also showcase how the
0
4
11
Here we show how the voxel grid function from the Ivy Vision library can be used in a TensorFlow model.
The same is true for PyTorch, Jax, and MXNet models.
Ivy and the Ivy libraries are not replacements for existing frameworks, they are supplementary!
1
0
10
Looking forward to presenting in 3 hours! We'll be explaining how to route to different LLMs for better performance 🔀. The event is open to all, so feel free to join! Thanks to Decoding Data Science for hosting us ❤️ See you there - happy prompting✌
0
2
11
🚀 New Blog Post Alert! 🚀
Discover how Ivy is solving the Framework Fragmentation Challenge in Machine Learning! 🤖🔗
Check out our latest blog post on seamless collaboration and code interoperability across TensorFlow, PyTorch, JAX, and more 👉
0
3
9
A leap forward in consistency 🔥kudos to the
@StabilityAI
team for pushing the limits of image gen with their new multimodality-infused architecture + repurposing of rectified flows 🌊 Exciting times!
In case you missed it, diffusion models can spell now! We released the Stable Diffusion 3 research paper yesterday, outlining the underlying technology powering the model.
Read the paper here:
22
59
403
0
2
10
New Models Have Landed on the Model Hub! 🦙🔥
⚡Llama 3 70b
⚡Llama 3 8b
⚡Mistral 8x22b
Check out the thread 👇
1
4
10
Thanks for the shout-out and for the contribution Rohan!
As mentioned, the application process to our open roles include a short paid task to our open source repo!
For anyone interested ➡️
Ever heard of getting paid to solve an assessment task for an interview? 🧐
Well here comes
@letsunifyai
who pays 25$ on upwork to complete the interview assessment task 🤑
How cool is that 🤩
#opensource
#Job
#interview
0
3
4
2
1
10
Really happy to be working alongside you Aman! 🙂
A year ago, I started my ML journey with TensorFlow only to realise a few months back, that I have to start over with Torch to collaborate with the community.
I recently discovered
@letsunifyai
, and it just blew my mind! I will be working on the codebase alongside
@DanielLenton1
0
0
5
0
0
8
This work was a collaborative effort:
@DanielLenton1
@PardoFab
@fabianfalck
@stepjamUK
@ronnieclark__
We welcome contributions, and encourage others to join the Ivy community by using Ivy in their own projects, and writing their own Ivy functions for inter-framework sharing!
2
0
8
🚀🚀🚀
8.3K ⭐️ Ivy is an unified Machine Learning framework.
@letsunifyai
#starhistory
#GitHub
#OpenSource
via
@StarHistoryHQ
0
0
1
0
4
8
We are really excited to welcome
@sh_reya
from
@Berkeley_EECS
, who co-authored the paper "Aligning LLM-Assisted Evaluation of LLM Outputs with Human Preferences". Happening Tomorrow! 🤩
RSVP 👉
The research introduces EvalGen, an interface that provides
0
3
9
Thank you very much for the contribution Anurag! 😊
Best of luck with the rest of your application! 🧑💻
Completed😍🥰😎 my UpWork milestone and contributed to
@letsunifyai
Thanks,
@DanielLenton1
👊for this great opportunity.
Looking forward to working with
@letsunifyai
Team
1
1
4
0
0
7
We're thrilled to announce that
@Vapi_AI
will be joining us for our weekly Webinar Series tomorrow! (Wednesday) 🤩
RSVP here:
Join us as we welcome Sahil Suman, Solution Engineer at Vapi AI, to the session. Discover how VAPI enables the quick setup of
0
3
8